Operating system preferences survey

My first home computer was a Sinclair ZX-80 which was interesting and the only one that fit my married enlisted home budget. Though my first computer was the university's IBM S/370. As a student reading IBM S/370 hex dumps, I could read machine code and EBCDIC character codes fluently by sight, but not Intel '86 machine code and in nearly four decades of working as a software engineer I never did pick up more than a couple dozen ASCII character codes, if even that many.My second home computer was a TI-99A with the peripheral expansion box which I got after starting my civilian career. It had MultiMate, which is how I learned how to use a spreadsheet application. When I upgraded to an XT clone (ran at Norton Factor 2) the difference was night and day. On the TI-99, you needed special software to examine memory and otherwise explore the computer, but on a PC the entire system was completely open to you. I was even able to identify and analyze my first Trojan malware, a downloadable graphic program to simulate a Star Trek star field view screen but behind the scenes it looked for the user and password files on a BBS system -- I became curious when it kept hitting on the hard drive, so I used debug to disassemble the program and figure out what was going on.I was also disappointed with the TI-99 when I learned how it actually worked. The advertising for it touted its microprocessor, the TMS9900, being a 16-bit device, but I could not find where it was even used. Instead, everything, including your BASIC programs, were run by the graphics processor. The only code that was ever run by the TMS9900 would be assembly programs that you had written. At least as far as I could see.BTW, at the time (1983-1984), my first civilian project was embedded programming for a training device that used the TMS9900, which was chosen in the original designs in 1977 because that was then the only existing 16-bit microprocessor. We did our development work on TI-990 minicomputers. That was part of why I had chosen the TI-99A when it was time to upgrade from the Sinclair.My TI-99 and expansion box are now stored in the garage. Does anyone know of any hobbyists or museums who might be interested in it? Decades ago PBS had a few shows done by a computer magazine columnist or the like who went through the history of the computer industry starting with the late 70's -- the first one was titled something like "Triumph of the Nerds: Accidental Empires." Lots of interesting information.In a later show, he interviewed the guys who developed VisiCalc, which was called the "killer app" that sold businesses on the Apple II. They were two hippie-type students, one a computer science major and the other an accounting major. In my 1970 Accounting 101 class, we were taught about actual spreadsheets which were basically large sheets ruled with t-journal columns that you would use as scratch paper for preparing your report, after which you would use the spreadsheet to stuff your flue (according to my teacher). They worked out a way to automate a spreadsheet which became VisiCalc. They couldn't sell it, but a salesman bought the whole thing from them for a couple hundred dollars and then turned it into a fortune.At the end of the interview when asked how they felt about how everything had gone down, they smiled and replied in hippie fashion (from memory): "Hey, we wanted to change the world. We did that."

We would never think of ASCII codes in decimal, but rather in hex since that's what we would always see in hex dumps. OK, I almost seem to recall that the chr() function in BASIC would use the decimal ASCII value, but in C we would just perform arithmetic on characters (eg, to convert a character for a digit into binary, we'd write something like n = ch - '0';, which would subtract the ASCII for zero ('0') from the character).'*' == 42 == 0x2A, but 'B' == 0x42 == 66Thinking of ASCII in hex is actually more useful, since the base values are consistent: See? The ranges for digits, capital letters, and lower-case letters are more consistent and easier to remember in hexadecimal than in decimal.EBCDIC (Extended Binary Coded Decimal Information Code) also has its own logic in hex. The two 4-bit nybbles that make up the 8-bit byte (each nybble represented by a hex digit ranging from 0 to 9, A-F) can be treated as if they were a Hollerith code (which I personally think EBCDIC was based on) with the higher order hex digit being the zone and the lower one the numeric. As in Hollerith, the numerics for letters were divided up into three groups based on the Navy DP mnemonic, "A J Squared away", with each group having a different zone (12, 11, 0 on a punch card); eg: As you can see, the digits are easy to remember and to read. There's no value for A1 or E1 since it follows Hollerith code in which punching two holes one right on top of another (eg, 0 and 1) might threaten the physical integrity of the card (I would guess). And as with ASCII, punctuation marks and control characters go elsewhere with little sense of a pattern. My TI-58 was my first programmable. Actually, my first programmable was an Hp 16C, their reverse-Polish programmer's calculator. It was programmable, but I never used it for that purpose, whereas I did play with programming the TI-58. Our first numeric methods homework was the problem of the infinite series of 1/x converging in the math, but not when you program it. So I programmed that into my calculator and let it run the entire weekend to verify that.When I first heard of scientific calculators in 1971, they cost about $250. Of course, the first 4-bangers cost $300 back in 1968, but then rapidly became so cheap that they were literally giving them away. I bought my first scientific calculator in 1973 for a reasonable price (it was in W. Germany, so I forget how many Mark it was) -- I forget the brand name, only that it was made in Cupertino, Calif.Then I had bought a TI-35 in the exchange in 1977 at a very good price ($30, I think). I'd gone through a few calculators, including one that was solar powered and had mixed-fraction functions. In the past year, I pulled my HP-20S programmable out of a briefcase (I had bought it in the late 90's) and I've been using it while writing my page on getting a feel for metric measurements. Timed delay loops. At one company we had sensors and control devices that communicated serially with RS232. But the EE decided to be extra cheap and didn't supply the board with a UART to handle the signal. So instead he implemented the RS232 signal processing in the microprocessor's code. Basically some delay loops, but mainly we had to ensure that each path in the code took the same number of machine cycles. Good thing it was a simple processor (8048, what IBM decided to use in the PC's keyboard).

One job I had was to support a computerized greenhouse control system. Most people might think that almost silly, but it is a most serious industry. Especially regarding when certain flowers can come to market. Most extremely important.A few times I had to go to a major customer's site with a then "portable" but rather what I called a "luggable." The unit was the size and about twice the weight of a piece of luggage. I think it was a ComPaq, complete with a built-in CRT (cathrode-ray-tube display). Whether it had an XT hard drive, I do not remember.That "portable", I would refer to as a "luggage". Something that was like a piece of luggaqe.

Pretty much it, except ours was an XT so it had one floppy drive and a hard drive (10 MB, I think). I didn't have to find desk space for it, but instead just stood it up on one end next to my chair (the keyboard was on a cord).At Hughes Aircraft (c. 1986) we did most of our work on a VAX network, but we also used first-gen Macs to make presentation slides, used Paint to combine text and graphics. When Windows 1 came out, I bought it for Paint, but it didn't support my printer. For me it wasn't until Windows 3.0 that it became useful.The Macs were diskette based, so we each got a system diskette and a data diskette. I labeled mine with my VAX username, dwise. When I got a second data diskette, I labeled it dwise2 and then added a "1" to the first one's label. Then one day a co-worker saw "dwise1" and laughed, so I sounded it out for the first time. When I got my AOL account I needed to come up with a screenname for the first time, so I used that. And that is the origin story of my username. When I was asking Kent Hovind for information about his solar-mass-loss claim, he tried twice to pick a fight with me over my username in order to avoid answering a simple question. Twice!Also at Hughes Aircraft I saw my first IBM PC up close. It was a DEC Rainbow, a clone of the original IBM PC (ie, no hard drive, just two floppy drives). It came with two or three OSes: MS-DOS, CP/M, and I forget what the third was (or even whether there was a third). In the user's manual instructions for booting it up, it even said "Please disregard the loud grinding noises you hear."

Around 1986 I upgraded from my TI-99A to an XT clone by Zenith and Heathkit. Worked for years until memory went bad so I donated it to Goodwill.It had a color CGA video board, 10 MB hard drive, and ran at Norton Factor 2 (ie, it was clocked at 8 MHz, twice the speed of a True Blue). And it ran MS-DOS 3.Later I came across other Zenith PCs, but they seemed to be older models. Plus their OS was Z-DOS.

As I recall, mine had a Zenith name plate, but then I would see Heathkit on the title page of the manuals.I resorted to consulting with Wikipedia a few hours ago: Zenith Data Systems

quote:

---

Zenith Data Systems (ZDS) was a division of Zenith Electronics founded in 1979 after Zenith acquired Heathkit, which had entered the personal computer market in 1977. Headquartered in Benton Harbor, Michigan, Zenith sold personal computers under both the Heath/Zenith and Zenith Data Systems names.

---

So then basically ZDS was Heathkit or at least Heathkit's personal computer operations. As I recall, the 8086 was the 16-bit version of the 8088, but since the original PC and the XT were 8-bit designs, they only used the 8088. The next PC, the AT, was based on the 80286.In 1985, I worked on a project that used a 8086, as I recall. And then starting in the late 90's our projects used the 80186.

One assignment in one of my classes (I think it was a EE class) required me to use a program on a PDP-8. As I seem to recall, it was parked in a hallway next to a Teletype Model 33 terminal (we all encountered one at one time or another; you can see one in the background three minutes into The Big Short).It was connected to the PDP-8 through a 110 baud serial port. Not only could I type faster than it could handle (and I'm not a fast typist), but I had to pause often to let it catch up.Edited by dwise1, : added qs block to provide context

We had those and they did have a certain coolness factor. The unit was a keyboard and dot-matrix printer in one. It stood on a stand under which there would be a box of 132-column fanfold computer printer paper that fed through the unit from below.It was featured on the July 1976 cover of BYTE magazine which showed Thomas Jefferson writing the Declaration of Independence on a DECwriter: Redirect Notice (scroll down to third image -- could someone able to post images here please do so?)In 1977, our university had an IBM S/370. Lower division students were required to use punch cards, but upper division students could get a VSPC account (Virtual Space Personal Programming). At the time, hardly anybody had a terminal at home, so we had to go to the computer center to log in. There were six DECwriters, each with a telephone and an acoustically coupled modem. We had a few 300 baud lines and several 110 baud lines; we were almost willing to kill to get a 300 baud line. I was lucky to have duty until midnight, so I would get to the computer center at about 0030 when the computer center was much less crowded than during the daytime and I would rarely have to wait for a terminal.Later we got a few ADM terminals with dedicated 300 baud lines and another year or two later a couple other terminals on 1800 baud lines. I preferred the 300 baud terminals because I would read the entire file listing as it scrolled by, but I couldn't do that at 1800 baud.We used VSPC mainly to write our programs (I think it used a line editor) and then to submit our job. VSPC also provided a version of BASIC, so I taught myself BASIC using David Ahl's computer games books as a model. Running Hamurabi and having the plague hit every single year made me realize that the RND function was different in VSBASIC than in Microsoft BASIC which the books were written in.FWIW, the Google image collection of BYTE magazine cover art is at byte magazine covers - Google Search .Share and enjoy!Edited by dwise1, : added having duty until midnight

Edited by dwise1, : added request to have image posted