Vintage Programming Languages

[ljones0]

Slightly OT I guess, but you can still get hold of copies of BBC basic - well - something extremely similar less graphics for modern PCs btw. If you go to this ( http://jaguar.orpheusweb.co.uk/branpage.html ) site you'll find a version of BBC Basic called "brandy basic" which will run on different OSes, such as linux (I use this) and dos.

ljones

[arjoll]

Wow, another thread that I hadn't spotted first (or second) time around

My first tinkering was Applesoft BASIC on an Apple II+, probably for 1/2 hour or so until the guy at the trade show demonstrating this flash new machine got annoyed

My first actual programming was probably SORD CBASIC on an M23 (our first 'home' computer - actually one dad bought to write survey calculations on, the same at the local council where he was County Surveyor) followed closely by BASIC on a ZX81 (my first computer), then SORD BASIC2, a 'structured basic' and back to Applesoft when school got their first Apple //es.

From then it jumped to SORD BASIC-68K when dad got an M68, and a little bit of playing with the version of C which came with CP/M 68K on the same machine (pre-ANSI and no floating point).

Then in 1988 it was over to PCs when SORD were taken over and shut down. GW-BASIC at school, QuickC at home, Microsoft C 6 once it was working.

Since then I haven't really moved on much I still have QuickC on my laptop, more for nostalgia than anything else. Microsoft C won't run under xp; haven't tried QuickC under Vista yet. 'Real' programming is either done with VBA for automating spreadsheets, writing custom functions etc, or MaxBasic (VB clone) and SQL to customise and automate the Accredo business software I support.

I have VC++, VC#, VJ# and VB express editions all downloaded from Microsoft and burned to CD but just haven't had the time to play....

[Paul Stenning]

This quote in post #1 was mine.

My programming is still mainly limited to BASIC, and I'm trying to get to grips with VisualBASIC 6. I have messed around with Perl a bit (out of necessity), and also used LabVIEW, Clipper (dBase language) and Forth where I used to work.

Although that was from 2004, it hasn't really changed much. I started off using BASIC on various home computers, a Sinclair ZX81 (OK for what it was and fascinating at the time as the first computer I had used), then a Commodore 64 (a big disappointment, fine if you wanted to play games, useless for anything else) then a BBC Model B which was a birthday present from my parents. For the following Christmas I was given a disk drive (and the DFS kit for it) and somewhere along the way a second hand Epson dot matrix printer was "acquired" (probably a cast-off where dad worked).

The BBC was an excellent machine for BASIC programming, and had an I/O port which meant I could combine my interest in electronics with my new-found interest in computers and programming. I also learnt something about wordprocesors, spreadsheets and flat file databases with this (due to obtaining copies of a couple of ROM-based office packages). I had this for several years, until I obtained my first PC.

This first PC was a 286 machine made by "Walters International" which was a cast-off from where my dad then worked. It was an IBM AT clone in a case nearly as big. It had been dropped by a carrier and did not work, even though someone had apparently tried to fix it. The problem was that a ribbon cable to the hard drive had presumably fallen off and been plugged on with one of the header pins bent. With that corrected it booted OK. The 40MB MFM hard drive was fine, though it had a few more bad sectors than listed on the defect label.

It came with a Taxan EGA monitor that had horrendous geometry problems. I wrote to Taxan asking how to obtain a service manual, and they sent me a photocopy of the circuit diagram, board layouts, setup info and common faults - for free. The problem was the first one on the common faults list - two electrolytics in the power supply (some things never change!).

Walters International also sent me a copy of the setup/utility disk (for configuring the drive controller, hard drive low level format etc) for free when I wrote to them. Companies seemed to be a lot more helpful then.

This was when 386 machines were starting to appear, and again it did me for a few years. In due course it inherited a bigger hard drive and RLL controller, around 180MB I think (again from dad's work).

Of course, PCs in those days didn't really have a BASIC programming language. Mine ran Compaq DOS 3.31 (I never had the original disks...) as it was then the only DOS to support partitions over 32MB, and BASICA was limited to put it mildly. At my work we had obtained Microsoft QuickBASIC 4.5 so I made a copy of that. We also had the mouse libraries with an early Microsoft Mouse (later they sold them separately). so I had proper DOS based programming on a PC. Excellent! A cut-down version of QuickBASIC (compiler removed) became QBASIC which appeared with MS-DOS 5.

At work I did some stuff with Forth (interesting, if a bit odd), Clipper (horrible) and even touched on a bit of TurboPascal (didn't do enough to form an opinion). The forth was on embedded stuff running Z80 based hardware. We used a product from MPE - ProForth I think.

I have also used LabVIEW quite a lot a few years ago. Although we use it in my current job, the program has moved on a lot and I have forgotten most of what I learnt, so someone else does that now.

More recently I have done a reasonable amount with Visual Basic 6. The only thing that has seen the light of day generally is the menu program on my CD-ROMs. I have done a couple of things at my current job with it too (it's good for putting together a front end for hardware which comes with a bunch of ActiveX controls, as long as you don't need great performance).

Its main use now is for manipulating data for importing and exporting to/from the eCommerce sites I run and their MySQL databases. I could probably write most of this in PHP and manipulate the databases directly, but I am only just starting to learn PHP. I know VB6 reasonably well, so it's easy to knock up a little program to do what I need to do in a few minutes. I can have it manipulate the text files I have exported or created, creating new ones to import, and not actually touch the database until I am happy with the import file I have created.

I suppose this could also be counted as programming... I have recently spent a few months at work writing PLC code using "ladder logic". The first rule when using ladder logic after using a proper programming language, I found, was to try to engage the brain into a lower gear! It really is dumbed-down programming for electricians (at least the Siemens implementation of it is)....

I have done a bit of stuff with Perl, mostly trying to modify the previous eCommerce script we used (an open-source thing called PerlShop). If you want to slow your web server down, just add a few Perl scripts!!

The current eCommerce software is JShop which is written in PHP and uses MySQL databases. I cannot modify the main code (well I could, but it would mean I couldn't install updates), but PHP can be included in the templates which is useful on occasion. This is why I am gradually learning bits of PHP.

[Dave Moll]

In case it may be of interest, I shall add my tale, beginning with my first introduction to the world of computers when I was in the fifth and sixth form (they're called something-or-other grade these days, aren't they?) at the tail-end of the 1960s. The school didn't have its own computer, but we were allowed to have our attempts run on the IBM 1800 belonging to the National Institute of Oceanography next door to the school. Mostly, these were written in FORTRAN IV, although this was also my first introduction to the use of assembler languages. Programs were submitted on decks punched cards, some of which I still have!

On leaving school, I got a job as a trainee computer programmer, working mostly in PL/I and assembler language on IBM/360 and IBM/370 mainfraime computers. I also did a small amount of programing in COBOL. Towards the end of my time as a computer professional, I had my first brush with the IBM PC and BASIC (which seemed very aptly named!) and (yet another) assembly language. It was also around this time that I purchased my first computer - a BBC Micro.

The BASIC provided with the BBC seemed to have far more functionality than the very basic BASIC I had met on the PCs at work, but even better was that it included an assembler language which enabled me to create free-standing executable code rather than the interpreted BASIC.

When I later acquired my very own PC (an Amstrad 1640), I found myself learning yet another assembly language. I then acquired a copy of Zortech C (a clone of Borland C - in fact clearly a name-change had been forced, as it was previously known as Zorland!).

edit: I forgot to mention the Amstrad PCW, which I programmed in Z80 assembler and Arnor C.

Other languages with which I have had glancing acquaintance include APL, Forth, Lisp, Algol and Pascal, but can't say that I recall anything from these.

Modern computer languages seem to have headed off in a rather different direction from my path in life - ho hum!

[paul.brett]

I started with BASIC on a ZX-81 that was delivered without a manual, but I got "Hello World!" working anyway.

Them I progressed through various other 8-bit machines and their various flavours of BASIC, until I got my first PC, and tried my hand at C. I never really got on with that.

I remember playing a tank battle simulation game on the PC called Omega, where the intelligence of the tanks had to be programmed. The idea was to go up against either AI or other people’s designs. This was certainly the weirdest programming language I ever used.

My first commercial exposure to programming was a modification of somebody else’s code on a Commodore PET that was running a postcode training program written in BASIC.

I thought Turbo Basic was great, and I actually wrote some commercial programs using it. I wonder if someone somewhere is still using them?

I have dabbled with various higher level languages at times, such as application macros (I built HUGE SuperCalc 4 macros when I worked at the Post Office)

Moving on, I have played with shell scripting on Unix and I have coded some very elegant solutions with DOS batch files as well.

I have tried to get into Perl and Java, but nothing more than tweaking other people’s work really.

Day to day tasks usually end up being written in BASIC still. All my current web site scripts (stuff that needs to be real-time anyway) are done in BASIC and compiled using Rapid-Q. All the non-real-time stuff is done using IBM Websphere TX.

Paul.

[dsergeant]

Originally Posted by ljones0

Slightly OT I guess, but you can still get hold of copies of BBC basic - well - something extremely similar less graphics for modern PCs btw. If you go to this ( http://jaguar.orpheusweb.co.uk/branpage.html ) site you'll find a version of BBC Basic called "brandy basic" which will run on different OSes, such as linux (I use this) and dos.

ljones

Even better is BBC Basic for Windows (http://www.rtrussell.co.uk/) - today far more advanced than the old Beeb Basic with full Windows API interface but will still run programs in the original code provided they don't use 6502 machine code segments. Not free, apart from a limited evaluation version, but worth every penny.

I started programming in Ferranti Argus machine code in 1968, long long before the days of the model B. None of these fancy compilers in those days. Was taught Fortran at uni, where they had written their own version Sofor (Southampton Fortran) which we prepared on punched cards. Later I was very adept with the Ferranti F1600 range machine code and paper tape.

Dave

[Dave Moll]

Originally Posted by dsergeant

Even better is BBC Basic for Windows (http://www.rtrussell.co.uk/)

I meant to mention in my post above that I do occasionally use that. In fact I was going to post a link to it yesterday, but the site had disappeared. I'm glad to find that it was a temporary fault rather than betokening the demise of the site (or, worse, of Mr Russell himself).

[rovereab]

I first programmed in City and Guilds mnemonic code in 1971 at college which was followed by Fortran IV. We used the Borough Treasurer's ICL 1902T mainframe and also a Honeywell timesharing service using ASR33 termiprinters as our interface. Aaah the joy of paper tape!

I then became a trainee programmer in industry and programmed Honeywell 2000 series mainframes in Cobol D and the assembler language called Easycoder (what a misnomer lol).

Regards.

Eamonn

[paulsherwin]

Originally Posted by Paul Stenning

Of course, PCs in those days didn't really have a BASIC programming language.

ISTR that the original IBM-PC had a very basic (sic) BASIC interpreter in ROM. The BIOS would fall back to this if all other booting options failed. IBM were very unsure where the IBM PC would finish up in the market, and engineered it so that it could be used without any disk drives at all, just like a ZX81

This was the case for all 8088 IBM PCs and the BASIC prompt caused lots of confusion in the office environments where it was used when it appeared.

My first programming language was Snobol4 - what's the betting I'm the only forum member to have written lots of code in that: http://en.wikipedia.org/wiki/SNOBOL

Paul

[Mike Phelan]

That's correct, Paul. The first PCs we had were the ones with a dual floppy drive (5150?) and booted up with the ROM BASICA. The next ones with a 10mb HD (5160?), and we thought "look at all this space - we'll never use it up!"

After the dual boot DEC Rainbow (8086 (DOS) + Z80 (CP/M)) it was a leap forward.
The DEC also had a BASIC in ROM, but it was even more basic than the IBM one - you had to select which line and character offset you were editing!

We used to moan about EDLIN - ED, the CP/M version was even worse.

Snobol4 - had a play with it. There was a freeware disk with some sample programs - random spoof company report and "insults2" which generated random pages of insults based on arrays of words that ran if any key was pressed; there was a bogus C:\> in the screen! Totally useless but fun!

[arjoll]

Paul Stenning's comment on Clipper reminded me of dark days - I have programmed in dBase IV which comes in useful when I strike Visual FoxPro on client sites Have also played with PHP, but it takes me about 20x longer to do something than it should do.

I had to use ED under CP/M-68K on dad's SORD M68. When we switched to PC I was spoiled by gradually better editors (the one in XtreeGold was quite good) but then thrown back into line editors when I started at Ernst & Young and had to use EDLIN.

In terms of actual computers, my first was a ZX81 which I blew up, then a second ZX81 I bought for $20 and sold for $20 a few years later (now they are >$100 ). My first PC was an "Ultra Turbo 286" (10MHz/1MB/20MB MFM). After working for a few months at my first job (accounting trainee) I spent $1,400 on a 120MB hard drive and a card for the new fangled IDE interface, only to find I also needed a new BIOS. The company happily sent down a couple of EPROMS which, surprisingly enough, worked on my motherboard. I remember buying a 2MB memory card (full length, 16 bit) from Otago University's Computer Services Centre when I was at uni, the 'click' on boot slowed down once it hit 640kb as my memory slowed from 10MHz to 8MHz.

Regarding BASICA, I remember the "Exzel XT" computers that replaced the Apple //e's at school in my 6th form (year 12) year. They had what was probably a dodgy copy of DOS3.3 (brand new at the time) which had both BASICA and GWBASIC. BASICA, of course, crashed the machines as they didn't have an IBM BIOS. Lovely computers, 30% failure rate, and no two machines with the same parts!

The directors of Exzel skipped the country about 6 months later after being sued by (and loosing to) IBM NZ for copyright infringement.

[OldBoffin]

Hi all,
This is too new for me! I started on a Hollerith Punched card machine in 1953 to get farmers paid for the weight of sugar beet going onto the factory ! Ended up with Patenting the Optical Encoder attached to a weighing scale pointer shaft, with Gray code out!
Then onto analogue computers using Magslips and Synchros with resolvers and Ipots to compute firing angles for warships!
Then Lab built nest of DC amplifiers to simulate and design networks for industrial servo systemsin such places as steel works and printing presses.
At last a proper computer - a Mathertron - with keyboard input, paper tape output and branching routines! WoW!
Next came "On Line" terminals to the USA, using teleprinters and tape punches in my lab - you joined the queue to do your work!
Then PDP8 with Forth etc actually in the Lab - it was all go in the 1960s
Then more work for the MOD - they wanted digital input from mechanical systems but would not use my Optical encoder because lamps could fail under the shock of guns firing ( LEDs hadn't been invented then)! So I made a Non-contacting electromagnetic encoder with BCD output.
Back to Analogue at 400Hz for the Admirals next with precision summing amplifiers ( 60dB gain with stability with 60dB feedback for unity gain if required) linling resolvers and other Synchro devices with a Digital to Synchro interface we built for them.
Then Apple computers appeared with Visicalc ( remember that? the first spreadsheet), followed in 1982 with BBC model B - by now I was Manager of an IT Centre and had 20 Beebs networked!
Soon supplemented by Torch machines ( a BBC with Z80 second processor)with networking and modem built in! Ran all BBC BASIC, also Wordstar, Calcstar,Datastar and Torch Finacial Accounts_ a real easy to use office machine. Also had Torch Mail between Cambridge and any owner in the UK =- all this in 1984 (way before the PC era!).
I learnt dBase 2 on this machine and still it together with three machines here if any one interested.
Torch at Cambridge soon added a third processor - a Motorola 68000 so we ran UNIX on our network. I spent many evenings with my assistant manager learning to use UNIX!
But now in 1986 the IBM PC appearedwith more packages to learn - dBASE3 followed by dBASE4 (Teaching packages available if any interest), also Lotus123 ( packages available ) Samna and Pegasus were next to try! Never a dull moment!
I retired in 1989 with an ACORN RISCOS machine - beats any PC and still in use today on my network here, with PCB design software which I used for the Design IT Teletext adapter board, and some EtherLan cards. But had to use a PC to learn Visual Basic for making analogue interfaces to a PC - quite easy when you get the hang of it!
I now have to dispose of all the equipment and software before I forget whats its all about! My RAM must be getting exhausted!
OldBoffin.

[Andy Day]

All this talk of high level programming languages gives me the horrors! As a dyed in the wool hardware man I got my head around machine code, in order to debug the hardware. It's fascinating to me to see the relationship between the code and the electronics, how at a higher level each bit may operate a certain function. After that, coming to terms with the microcode that interpreted the machine code was a wonder, and the absolute key to hardware faultfinding. But then, even though this was all designed 40 odd and more years ago, similar microcode is still deep inside the most modern of processors, and the only way these can be accessed is via a translator with a suitable output. (Which is what a language is, a translator from language to machine code)! .

[Skywave]

Originally Posted by evingar

I much prefer hardware, but from time to time I am forced into blunder around with software.

I have tried my hand at C++ but after the elegance of Pascal, I find it a " messy " language.

Sentence 1: agreed.
Sentence 2: agreed.


Al / Skywave

[Skywave]

If it's not a daft Q., does writing batch files to run under DOS count in all this?

And does writing macros for "Supercalc" count? (Pun not intended! )

(The hours that I've spent on those two )

Al / Skywave

[Mike Phelan]

Al
I suppose they are all "languages"! As far as C, or C++ vs Pascal, I always think of Pascal not allowing to hang yourself, but with C, you can because you might just need to!

Everyone
I started with dBaseII running under CP/M and progressed to dBaseIII when the PC and DOS arrived.
After everyone started to moan about significant features (like arrays and a compiler) not being implemented, some of their tech guys jumped ship and created Clipper.
Despite the comments above, it was an absolute godsend - we could use existing tables, and the users could not get at the source code.

I researched 4GLs for Clipper to speed up production at work, looked at Genifer and UI Programmer (good in their own ways), and finally settled on Robert Bolton's marvellous Sycero dB.
You could create a useful Stock Control system in a couple of hours. The true test of a 4GL, I never needed to edit any Clipper source in 50+ applications.

Sadly, threw all of this in the bin a few months ago - could not even give any of it away.

For my sins, I still maintain a Visual Foxpro application that runs a large company. They could not find anyone else anywhere near!

[Steve_P]

Well it was fun for me too. Somewhere in the attic is a full set of commands and instructions for a Z80 processor chip. And a book about all the other chips that came with them.

Anyway, my school began with a book by ICL/CES which taught us all about stuff we'd never see, like OCR Readers, Card Readers, Writers and Verifiers. There were some given to school, and I got them running, only to find that their only use was to give kids a card each with their name on. And other things....

I did high and low level languages, dry running, flowcharts and even how to add and take away in Binary, Hex and Octal. Unfortunately, as an Electronics Freak even then, I came top in this and was treated accordingly, even worse was to come.

Here's one language that has not been mentioned - CESIL. Stands for Computer Educational Schools Instruction Language. Sort of Machine Code with words. Yuck!

Anyway, did Forth, Algol and a few others for fun at home, but the computer at home and the one at school were different. I had a ZX80 at home (The White Flimsy forerunner of the ZX81, and you could get an 8K Rom upgrade to ZX81 which I did, and made a switch arrangement so I could use both!)

At school, we had an 8K PET with a printer etc. There were games like Rhino and Nightmare Park for this, which I got into, printed off and transported onto ZX Spectrum later on. There was a bug in Nightmare Park, which I fixed!

As I said just now, there were lots of kids with other machines, and school said that the only correct form of basic was theirs. I had some rows with the teacher over this! Especially when I found out that they were teaching PET Basic (as they called it) wrong as well. I ended up doing some of the homework twice over, in a table, to show up the differences. This surprisingly did not go down well, and I got comments like 'Not on a PET' across them.

Once, in class, I gave an answer which the teacher was convinced was wrong, so, expecting me to fall flat on my face, I was allowed, with the whole class around me, to type it in. I did. The teacher didn't speak to me for a week...! Later that year, they couldn't get READ and DATA to work right. I knew what they were doing wrong, but kept my mouth shut, telling my friends and nobody else.

Then on to the BBC, where I learned BBC Basic as well as Machine Code, Forth and how to program a MIC system. But that's another story...

Cheers,

Steve P

[RobinBirch]

Ah CESIL I've come over all 1973 again. We did this briefly before moving on to SIMPLE.

I seem to recall that we wrote our programs on programming sheets which were then sent to the council offices in Haverfordwest and we then got the results back as a line print. Sometimes with red ink all over it from when the coders in the offices were bored and couldn't take punching in broken code any more.

Cheers

Robin

[GMB]

I started computing in my teens using Zebra-autocode (an early decimal machine) and quickly moved on to BASIC and FORTRAN-II
I then learned two machine codes, of the XDS-Sigma-5 and Elliot-803. The Elliot introduced me to Algol-60 too.

At Cambridge the computer science course mainly used ALGOL-W, BPCL and LISP as well as the IBM/370 and PDP8 machine code.
Back at work I encountered COBOL, ALGOL-68, FORTRAN-IV, SNOBOL, APL the survey processing language SPS.

The first major compiler that I wrote was for PASCAL - and I was one of the pioneers of the idea that PASCAL could be adapted for writing "system software" which up to that time had always been hand-written in machine code. My extended PASCAL became known as PASCAL-T and was compiling for Sigma5/7/9, DEC 10/20, IBM 370 and Interdata 50/74.

I studied MODULA, LIS and ADA which were other Pascal-like languages and for some reason the UK MOD decided to adopt CORAL while the US were still pinning down ADA and I nearly had to write a compiler for that.
For some reason that I forget I wrote a CESIL compiler and runtime, and also a LOGO system.
I looked at FORTH but found it too hard to write working programs, but liked the idea so developed a threaded language of my own that was more robust.

Then I got the chance to write a survey processing language called ABC, this time for the Mac. By this time I had learned several more machine codes for the new generation of microprocessors, M6800, Z80, 6502, 8080 then 8086 and M68000.
I briefly encountered SMALLTALK, SQUEAK and MIRANDA while looking for inspiration for NDP which was my first shot at a functional dataflow system for serious use. I also wrote a PostScript compiler.

Of course it's hard to avoid C and all its many derivatives and I even wrote a JavaScript system, but I really don't like their syntax and I consider C is just a way of generating bugs easily!

I think this qualifies me as a "vintage compiler writer" - and you can parse that either way!

[Terry Judkins]

Originally Posted by Andy Day

All this talk of high level programming languages gives me the horrors! As a dyed in the wool hardware man I got my head around machine code, in order to debug the hardware. It's fascinating to me to see the relationship between the code and the electronics, how at a higher level each bit may operate a certain function. After that, coming to terms with the microcode that interpreted the machine code was a wonder, and the absolute key to hardware faultfinding. But then, even though this was all designed 40 odd and more years ago, similar microcode is still deep inside the most modern of processors, and the only way these can be accessed is via a translator with a suitable output. (Which is what a language is, a translator from language to machine code)! .

I agree. As an IBM tech from 1963-1993 I was required to know the hardware and the microcode that manipulated it. Early on it was all proprietary stuff. The 2030 processor had microcode stored on IBM cards made from a special plastic with conductive pads where the holes should be. Punching or not punching out the pad created ones and zeros for the microcode words. The 2040 (designed in the UK BTW) had TROS modules (Transformer Read Only Srorage) that used inductive rather than capacitive coupling to form the ones and zeros. Later machines like the 3145 processor had microcode stored in volatile memory which had to be reloaded at each power on. I wrote some diagnostic routines for the 3145 which used a 32 bit microcode control word.
Later, in the 80s, the 3090 and 9021 processors used VM and REXX for firmware. There was a lower level microcode buried deep in the hardware which we were not privy to however.

After my retirement I made use of the VM and REXX knowledge to write a 70,000 clause application while working for a company with a contract with IBM. REXX was a high level language similar to PERL It could be run as an interpreted script or compiled and run as a standard program. REXX was very flexible and powerful. It is a shame that it has gone by the wayside with the rise of the PC and the fall of the mainframe. There was an effort to transition REXX to the PC but it never caught on as far as I know.