SoylentNews is people
SoylentNews
10 Most(ly dead) Influential Programming Languages:
The other day I read 20 most significant programming languages in history, a "preposterous table I just made up." He certainly got preposterous right: he lists Go as "most significant" but not ALGOL, Smalltalk, or ML. He also leaves off Pascal because it's "mostly dead". Preposterous! That defeats the whole point of what "significant in history" means.
So let's talk about some "mostly dead" languages and why they matter so much.
Disclaimer: Yeah not all of these are dead and not all of these are forgotten. Like most people have heard of Smalltalk, right? Also there's probably like a billion mistakes in this, because when you're doing a survey of 60 years of computing history you're gonna get some things wrong. Feel free to yell at me if you see anything!
Disclaimer 2: Yeah I know some of these are "first to invent" and others are "first to popularize". History is complicated!
<no-sarcasm>
If there were one perfect language we would all be using it already.
</no-sarcasm>
Recently:
(2020-03-11) Top 7 Dying Programming Languages to Avoid Studying in 2019-2020
(Score: 3, Disagree) by istartedi on Friday March 27 2020, @06:41PM (10 children)
If you've never heard of MUMPS (I hadn't until just a few years ago) google around and read some stories. I think it's influential as an example of what *not* to do, and how you can get stuck with something awful because of sunk costs. It would be totally dead, except that it's too hard to kill. COBOL is a lot like that, but I'm given to understand that COBOL is quite sane compared to MUMPS.
Appended to the end of comments you post. Max: 120 chars.
(Score: 2) by DannyB on Friday March 27 2020, @06:56PM
Very [wikipedia.org] Interesting. [uni.edu]
Interesting that a GPL version is available.
I didn't know that BASIC (once my favorite language) and MUMPS shared a common ancestor: JOSS.
Glad its not contagious. (but isn't the GPL a viral license?)
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 2) by maxwell demon on Friday March 27 2020, @06:59PM (7 children)
The name suggests that the authors of MUMPS never actually intended it to be used. I mean, what would be a more effective way to signal “stay away” than naming your language after an infectious disease? On the other hand, maybe they hoped it would go viral with that name.
Someone should invent a bad programming language and name it CORONA. :-)
The Tao of math: The numbers you can count are not the real numbers.
(Score: 4, Funny) by istartedi on Friday March 27 2020, @08:39PM (1 child)
Continuation Oriented Recursive Object Numerical Analyzer?
Features: No function is valid unless it uses call/cc in some way, and if you attempt to encapsulate call/cc in some other control structure, it's an error. All data structure definitions must be self-referencing, and any attempt to create a data structure that is non-numerical such as a "string" is detected by the compiler and is an error. If there are fewer than 2 monads per line of code, that's an error. Somewhere between 30 and 40% of functions must use point-free style (infix function composition with the dot operator) or the program won't compile. Must indent 5 spaces on each line, although two slashes denote indentation of blocks. Comments begin with a curly brace, and end with a right parenthesis.
Appended to the end of comments you post. Max: 120 chars.
(Score: 2) by All Your Lawn Are Belong To Us on Monday March 30 2020, @09:39PM
must also feature an lp0 printer on fire? error, feature no garbage collection so the main memory is slowly drowned with built in requests for more memory than required for any running function, and must periodically push random core dumps to the screen (Common Object Universal Glyph Html) that may cause the program to propagate to other machines.
This sig for rent.
(Score: 3, Funny) by DannyB on Friday March 27 2020, @09:32PM (4 children)
Someone should invent a bad alcoholic beverage and name it CORONA.
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 3, Funny) by MostCynical on Saturday March 28 2020, @12:21AM (3 children)
hey, cold, with lime or lemon, it is drinkable (yes, so is Fosters [wikipedia.org] and XXXX [wikipedia.org], so YMMV..
"I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
(Score: 1, Touché) by Anonymous Coward on Saturday March 28 2020, @02:03AM
Agree is a far better lemonade sub.
(Score: 2) by All Your Lawn Are Belong To Us on Monday March 30 2020, @09:40PM (1 child)
Fosters is drinkable? Coulda fooled me!
This sig for rent.
(Score: 2) by MostCynical on Tuesday March 31 2020, @12:37AM
these days, only by the British [theguardian.com] and not really very much in Australia [quora.com]
Although the stuff the brew in Manchester is very different to the stuff made in Australia.
"I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
(Score: 0) by Anonymous Coward on Friday March 27 2020, @09:04PM
Only when you include the sugar. Comparing vanilla COBOL to the sweetened syntax should make you cry.
(Score: 2) by JoeMerchant on Friday March 27 2020, @07:05PM (15 children)
Not invented here syndrome will forever stop us from using the "one perfect language" - even if it existed.
I've heard of a half dozen or so locally grown script languages that started as a simple exercise and evolved to the point that Python would have been a much better choice, but too late, everything is in the home grown script.
I nominate Fortran for highly influential. Straightforward, stuff starts happening more or less right away, not a lot of declaratory cruft, basis for BASIC and any number of other "simple introductory" languages. And, the supercomputing crowd can't seem to wean themselves from it...
🌻🌻 [google.com]
(Score: 0) by Anonymous Coward on Friday March 27 2020, @07:34PM (10 children)
I had to actually look at the article to confirm that FORTRAN was indeed not on that list, which is absolutely ludicrous.
(Score: 2, Interesting) by Anonymous Coward on Friday March 27 2020, @07:46PM
FORTRAN is still going, but it is alive only in a very particular niche: high performance numerical simulations.
I was surprised to learn that some old popular software (like text adventures, other stuff) from the 60s and early 70s were written in Fortran. Even 30 years ago, I thought it was purely used for number crunching! I never thought it would be used for text processing, but then, we used to process text using C, and that language doesn't even have a real string type. I guess you just use whatever language is commonly available more than anything else.
(Score: 2) by DannyB on Friday March 27 2020, @09:34PM (6 children)
The article points out that FORTRAN is not quite dead. Therefore it does not quality to be on the dead languages list.
But what about dead natural languages? Latin? French? Etc.
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 1) by pTamok on Saturday March 28 2020, @10:33AM (5 children)
Latin isn't dead. It is still, for example, the official language of the Holy See.
There are also plenty of books of historical interest written in Latin, for which a working fluency is necessary.
There are plenty of natural languages that are dead. No written records, last speakers dead.
(Score: 2) by maxwell demon on Saturday March 28 2020, @12:05PM (1 child)
Actually there are also dead languages with written records, but nobody being able to read those.
The Tao of math: The numbers you can count are not the real numbers.
(Score: 1) by pTamok on Sunday March 29 2020, @10:05AM
Aren't they zombie languages?
I thought of Linear-A, the Phaistos Disc, Rongorongo, and the Voynich manuscript: any others?
...to answer myself, of course there are: there's a Wikipedia list of undeciphered scripts [wikipedia.org].
(Score: 0) by Anonymous Coward on Sunday March 29 2020, @12:08PM
I was taught in high school that dead languages were defined as those that never change. No words added. No meanings change. There was no qualification of speaker existence or written texts. Since some of these programming languages are under active development, that definition does not apply. Of course this definition was in regard to spoken language.
Of course the definition could have changed. "Scholars" from generation to generation tend to change definitions of commonly accepted things because reasons. I guess you need to do something to justify your existence and get published.
(Score: 3, Touché) by DannyB on Monday March 30 2020, @02:15PM (1 child)
Okay. Maybe Latin is not really dead.
It is informative that you don't raise any objections that I mentioned French is a dead language not being used.
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 1) by pTamok on Monday March 30 2020, @09:11PM
I've moderated you "Touché" for that.
The useful working definition, that someone else has provided, of a dead language being one that does not change, obviously rules out French as being dead.
But you slipped it in nicely there. Please excuse the double entendre, which, curiously enough, is not a phrase used in French. In French, such word play is termed double sens [wikipedia.org]
(Score: 3, Interesting) by sjames on Saturday March 28 2020, @01:02AM
FORTRAN is alive and well for numerical simulations. It will probably be here a very long time. Part of it is that it is extremely difficult to validate a numerical model to any degree of certainty, so when you have code that works, you try to disturb it as little as possible. That's why even in software still under development, you still find big chunks of code in FORTRAN 77. It's not at all uncommon to find software that still refers to the input file as a 'deck' and individual lines in the file as 'cards'
Another reason is that by it's nature it's easier for compilers to optimize FORTRAN. No pointer alias problems, the language defines the size of variable types (where in C, until recently, the machine defined the types) etc. While compilers for other languages have caught up now, in the '90s and early 2000's, FORTRAN code simply compiled down to faster executables.
(Score: 1) by pTamok on Saturday March 28 2020, @10:27AM
FORTRAN is still going as a business language as well. Well, the DEC/VAX variant. I know of one major business that is using it, together with the DEC/VAX character-based screen-forms libraries to run applications. It works, the engineers like it, and the is no justification for spending money to replace it. Eventually it will die when the business decides to migrate the data into a newer application.
But yes, FORTRAN was used as the glue for business applications. You can write procedural logic in almost any language.
(Score: 0) by Anonymous Coward on Friday March 27 2020, @08:50PM
Not invented here syndrome will forever stop us from using the "one perfect language"
English is close enough for me.
For computers, you're supposed to speak binary.
(Score: 2, Interesting) by Anonymous Coward on Friday March 27 2020, @09:11PM (2 children)
Similarly, Lisp will never die. It is the de facto language for software development on new hardware developments. One of the first things they do is write a lisp cross-compiler for it. As the best example of that, all quantum computing code is written or transpiled to lisp.
(Score: 0) by Anonymous Coward on Friday March 27 2020, @10:45PM (1 child)
Doesn't match with my experience. I would say writing a simple C compiler for it comes very early. But who knows, maybe C is the first compiled language they create and LISP the first interpreted. I don't know.
(Score: 1, Interesting) by Anonymous Coward on Saturday March 28 2020, @12:18AM
C compilers, specifically cross-compilers do come very early. They are considered one of the first steps to commercial support because getting one allows everything that relies on C or the standard library of C to run on it. I'm talking about before that. They often use Lisp because it is fairly easy to turn a subset of Lisp into machine code. These basic structures allow for easy verification and testing of the integration of the hardware units during late development and validation.
(Score: 2) by pkrasimirov on Friday March 27 2020, @07:22PM
I guess it's not in the list because probably it is completely dead.
(Score: 1, Informative) by Anonymous Coward on Friday March 27 2020, @07:39PM
I nominate Ruby.
Truly a historic and highly influential language that everyone knows.
Also, C# 2.0
(Score: 2) by looorg on Friday March 27 2020, @07:48PM (1 child)
This again. I strongly suspect COBOL will outlive us all.
(Score: 5, Funny) by DannyB on Friday March 27 2020, @09:36PM
In about the year 9997, when the Y10K problem is right around the corner, and everything must be udpated for five digit years, there will suddenly be a lot of job openings for COBOL.
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 4, Interesting) by Samantha Wright on Friday March 27 2020, @08:11PM (2 children)
I... don't think the author should have written this article; it's rife with errors, distortions, and even a few pieces of outright misinformation. A PLT (programming language theory) specialist would've written a much shorter list, with little overlap. And an industry-centric languages list would be much longer, again with many omissions. Even a list of most influential languages (in a genealogical sense) would look quite a bit different from this one. It is regrettable that this is the article that made it to SoylentNews.
(Score: 3, Interesting) by RamiK on Friday March 27 2020, @08:56PM (1 child)
Honestly most language theorists have no clue what makes a language good. And who can blame them when most developers will opt to use the tool they're most familiar with rather than the tool most appropriate to do the job?
compiling...
(Score: 5, Insightful) by Anonymous Coward on Friday March 27 2020, @09:48PM
You are right, but...
the best tool is usually the one you know inside-out.
The ramp-up time and potential unknown traps of a language that is new to you usually outweigh any theoretical benefit of adopting a new language along with its supporting ecosystem.
I am assuming the language you already know is a general purpose language: probably not the best at anything, but also not the worst either.
(Score: 3, Interesting) by Anonymous Coward on Friday March 27 2020, @10:09PM (1 child)
Dartmouth DTSS BASIC was a compiler not an interpreter. You would input, edit, and save your source code using a line editor on your Teletype printing terminal. Lines starting with a digit were placed in the file being edited. Lines beginning with a letter were commands to the OS or to the editor. The RUN command would compile your source-code file, run the object code, and then discard the object code. There was a separate command to compile and keep the object code. I used DTSS BASIC from 1971 to 1973. See John Kemeny's 1971 book for details. Microcomputer BASIC's were interpreters to avoid needing a disk.
(Score: 2) by DannyB on Monday March 30 2020, @02:29PM
I used a minicomputer in college that you probably never heard of. (Harris S-100 (no relation to the microcomputer bus), and os was Super Vulcan)
Among various language compilers and interpreters, it had BASIC.
This BASIC was very complete. Much better than most microcomputers. It also was available in two forms:
1. A batch compiler (eg, like FORTRAN, etc)
2. An interactive compiler
The interactive compiler is interesting. It acted in every way like an interactive BASIC interpreter. But each line was compiled as you typed it. It kept the source representation as well as a compiled set of instructions for that line.
The performance was in all ways the same as the batch compiler.
I can only speculate about how this worked. I assume that when you typed RUN, it had to patch all the Jump instructions from all GOTO and IF/THEN statements. I suspect that the last instruction of every compiled line of code was a JUMP to either the interpreter's "next line" subroutine, or maybe directly to the front of the compiled code for the next line of code.
Every performance optimization is a grate wait lifted from my shoulders.
(Score: 1, Funny) by Anonymous Coward on Saturday March 28 2020, @06:21AM
Born in academia, it still serves academia.
(Score: 0) by Anonymous Coward on Saturday March 28 2020, @01:47PM
Sigh.
I work in a shop mostly written in PL/I. The core systems are running on IBM Mainframes. Some newer parts are written in Java 11, but we also 8, 7 and 6.
It's a heck of a job to modernize.
(Score: 2) by ChrisMaple on Sunday March 29 2020, @03:15AM
Honeywell had a PL/I compiler. I don't know when it came into existence, but I used it at MIT circa 1970.