Over on GQ, they’re trying to decide which programming language would be best for someone just starting in the field. Some of the languages they are sugesting I’ve never heard of, e.g., Haskell, F#, lua; others seem like the latest hot flash, e.g., Ruby, Python.
I took my first programming class in 1969 when COBOL and FORTRAN were all the rage, and since then I’ve seen languages come and go. One of the reasons I got out of the business was that I never felt I could develop a real sense of craftsmanship with any technology. By the time I felt I was really getting good with the tools at hand, it was time to move on to the latest greatest thing that would solve all of our problems - usually problems I didn’t even know I had.
That got me thinking about all the now useless languages that I had to learn to keep up with the market. For me that would be:
and operating systems like:
[ul]
[li]VAX[/li][li]DOS[/li][li]whatever the old IBM mainframe OS was called (JCL was the language I think)[/li][li]Univac was in there somewhere also[/li][li]lots of others I’m sure I’ve forgotten.[/li][/ul]
Well, that’s a good point. In 2004 I was looking for a job, and AC Nielsen was looking for somebody who knew FORTRAN. Languages are not quite dead for a long, long time.
(Incidentally, they didn’t bring me in for an interview because I was too senior - they said it was an entry level position. I wish I’d had the presence of mind to really make fun of them to their face for that, but I desperately needed a job, and don’t always think fast on my feet anyway).
Turbo Pascal - if it’s not dead by now, it should be.
I also used Clipper once upon a time way back when.
Back when I was working on OS/2 I did a lot of REXX and we also had something that was called VX-REXX, which was sort of like VB but with REXX underneath. We also worked with a data entry software package that had it’s own proprietary language which was like the bastard love child of COBOL and assembly. (I’m glad my OS/2 days are over!)
Atari 8-bit BASIC, BasicXE, Turbo BASIC, Action!, 6502 assembler, 8086 assembler, Logo, GFA BASIC, Clipper, dBase III. (The last two I did in college, along with 8086 assembler, and boy, did I hate me some Clipper and dBase III. I loved assembler coding, though dealing with segmented memory on the 8086 was a pain.)
I’ve also played with Bash/CShell/TSCH scripts and stuff, as well as IRC script. (Made me a bot for an IRC channel I opped.) These aren’t exactly dead though, but they’re relatively obscure by comparison.
The OS for 360s was OS/360, though many other variants got added. JCL was the job control language, which controlled the priority of your job, which resources it needed, etc. I only know it from a class as a part of our system programming class at MIT, which tried to bootstrap on us enough to be able to submit jobs.
I know lots of dead language.
BCPL - the precursor to C.
IBM 1620 assembler
PL/1.
Algol - though I never actually programmed in it
SNOBOL - ditto.
PDP-11 assembler
CDC Cyber Assembler - taught these last two.
Tutor - The PLATO language for lesson creation
If you call Pascal dead, know and taught that, and even hacked the Zurich compiler.
LGP21 machine language - and the assembler I wrote for the machine in high school.
And last but not least, the language I invented and implemented for my dissertation, which had a short life after it in a proposal RCA put together for the government.
I did some programming in MS-Basic, but the language I did some really complicated programming in was Forth. That is still the language I feel most comfortable with, but it was victim of its clergy who insisted that it stay primitive (no string handling, no floating or even fixed point arithmetic (only integer) and it died. To be sure you could build things that would carry out those procedures, but everyone rolled his own and no two were compatible.
I also regret the passing of RPN calculators. I never feel comfortable with algebraic calculators. Nowadays it would be entirely possible to build that would switch with a flick of a switch, but, AFAIK, no one has.
RPG - Report Program Generator. IBM mainframe column-based programming language. I only knew enough for maintenance of a small number of programs, and I made a very minor hack to make the compiler work with 3330-type disks (that REALLY dates me). I do NOT miss it. There’s now RPG IV, which I haven’t seen, but from reading wikipedia it probably doesn’t look a lot like the original.
APL - from the IBM 5100 desktop computer. Actually a very interesting language which had very powerful handling of arrays of any dimension. But programs could be hell to maintain.
I had the little ‘memory map’ book by Jim Butterfield listing the values stored at boot for every addressable memory location on that machine, and what that location was used for. Man, those were the days.
I’d disagree that COBOL is dead. There’s still a lot of business’ using older systems in COBOL. Even some of the canned software uses COBOL. People Soft used it up until about 6 years ago. Our shop converted from in house software to IA in the early 1990’s and IA was in COBOL. IA included a registration package, finance/payroll and HR.
A lot of COBOL people have retired. The ones that are left are in a great position to earn top pay. I have 24 years experience in COBOL.
Similarly, there’s no way that Fortran is dead (unless you’re counting one of the really old versions of it). My flatmate was taught Fortran as an undergraduate, and uses it for his research (Geophysics). It’s still the language of choice for numeric computing.
Indeed it isn’t - I’m retired, but still work part-time supporting old production FORTRAN programs.
In addition to many of the already-mentioned languages, I used several while at college in the 60s that probably never left campus - CUPL (Cornell University Programming Language) and ZAP (Zelkowitz’s Assembly Program, an IBM 360 assembly language interpreter used as a teaching tool).
But my best stories come from the days when I programmed business systems on tiny computers that were hand-built by an entrepenurial engineer (Microdata 4004 CPU, 8K total memory, 5MB hard disk drive, had a screen and keyboard but used a Teletype and its 10CPS paper tape drive as its programming I/O system). The only language provided for those computers was MicroData 4004 assembler, and the CPU didn’t have a Multiply function (I had to write a multiply subroutine using shift and add instructions). Ahh, those were the days - and I hope I never have to live them again.
Ten minutes of frantic googling indicates that Microdata apparently never made a CPU labelled “4004” - I’ve apparently misremembered that number (and possibly even the chip manufacturer) in my old age.
If anyone else remembers old CPUs, this was an 8-bit CPU that featured three registers - an 8-bit “general purpose” A register, an 8-bit “general purpose” B register, and a 16-bit X register that was used for indexing purposes only and had to be loaded from the A/B registers in two passes. Technically you could have 64K of memory attached to the CPU (because of that X register), but it could only run programs out of the first 8K, and the only way that a program in either 4K segment of that lower memory could access code in the other 4K segment was via a “Jump Other” instruction that would transfer control to an instruction in the “other” 4K segment of memory. The CPU had add and subtract functions, but did not have either a multiply or divide instruction. We used this CPU in the mid 70s, but it may have already been technically obsolete by then.
Not to get into a debate on whether they’re actually dead or just moribund, but here’s my list:
COBOL
FORTRAN
PL/I*
C
Pascal*
SNOBOL*
Motorolla 6800 Assembler*
CICS
Basic (The kind that came free with your MS-DOS operating system)
SQL Windows
RPG III*
PCS/ADS
There’s now RPG IV, which I haven’t seen, but from reading wikipedia it probably doesn’t look a lot like the original.