Most(ly dead) Influential Programming Languages

https://hillelwayne.com/post/influential-dead-languages/

by luu • 6 years ago

417 289 6 years ago

DylanSp• 6 years ago

Glad to see a mention of CLU; it pioneered a lot of ideas. Reading Joe Duffy's blogs about the Midori project (greenfield language/system design at Microsoft), he mentions CLU as an inspiration a number of times.

pjmlp21• 6 years ago

One can still play with CLU today,

http://www.pmg.lcs.mit.edu/CLU.html

I would already be more than happy if Go 2.0 generics would be CLU like, no need for something more fancy.

pjc503• 6 years ago

Despite having a CS degree I had never heard of CLU, which is a shame. Author is Barbara Liskov of the Liskov Substitution Principle, influential in early OO design.

dionys2• 6 years ago

Interesting to see Pascal on the list - it was the language of choice in my high school and my first programming language, back in 2010. I haven't touched it since, but I do remember it was pretty easy for me to pickup the syntax, especially compared to languages I would learn later in uni, like say C.

greencore2• 6 years ago

In Lithuania and perhaps in similar countries Pascal is still main programming language in classrooms. Also no other language has beaten Pascal in number of released books so it will stay for very long time.

non-entity• 6 years ago

I almost started a new retrocomputing-related project in pascal recently because FreePascal seems to be the only stable maintained toolchain that still can target 8086

dionys2• 6 years ago

I can confirm! My high school was in Slovakia and I know other people had the same experience with Pascal.

QuesnayJr5• 6 years ago

It took me years before I stopped accidentally typing ":=" in my C code.

FpUser4• 6 years ago

I switch quite often. Sometimes 3 languages a day: Delphi/Lazarus/FreePascal, C++ and Javascript so it now takes few minutes to adjust.

benibela• 6 years ago

I still write almost all my code in Pascal

It compiles very fast to native code, has properties with implicit setters/getters, null-safe strings with mutability xor aliasing, function and operator overloading.

It is really amazing

gwbas1c• 6 years ago

Wait, Pascal was in use in 2010? I thought it was a dead language when I used it in high school in 1997-1998.

kick4• 6 years ago

Pascal has been in an incredibly strange state of being both dead and alive simultaneously for the past thirty years.

...but yeah, it was definitely doing pretty good throughout the 1990s.

Tyr42• 6 years ago

I know Turing (which is pascal packed up with a graphics library and manual and editor all in one exe) was in use in Ontario Schools as late as 2010.

abetlen• 6 years ago

Weird right? I learned Pascal in highschool in Ontario around 2010 as well. I remember being really frustrated at the time because we weren't learning some modern language like Java or Python. Instead we were stuck with Turbo Pascal, writing in an IDE so old it only supported 8 character long filenames. Thank god we had an amazing teacher which ended up making it one of the best CS classes I've ever taken.

badsectoracula3• 6 years ago

Well, depending if you count Free Pascal and Delphi (aka Object Pascal) or only Wirth's original Pascal, it is still used today.

jodrellblank2• 6 years ago

Turkey bought Delphi licenses for a million students just a couple' months ago: https://news.ycombinator.com/item?id=22116275

bryanrasmussen2• 6 years ago

I linked to something on Erights.org a few days ago http://www.erights.org/

I think E was more influential than its distribution would suggest.

evgen4• 6 years ago

As cool as E was, it was almost entirely based on core ideas lifted from Joule and KeyKOS. I can't really see any influence beyond a "wouldn't it be great if language X had feature Y like you find in E".

bryanrasmussen2• 6 years ago

I was also thinking of the community, as they moved out into other languages and carried some ideas with them that they maybe got exposed to in the E community.

pjc503• 6 years ago

There are a few things present in COBOL that I would like to see lifted into more modern languages. One is the use of the full stop as an expression terminator, as in English, instead of the semicolon. Another is the whole "picture" mechanism for number formats.

More languages should also steal Verilog's use of _ as an optional digit separator. It's just arrived in C# 7.

gerikson3• 6 years ago

"_" has been an option to separate digits in Perl for ages. I use it all the time if the integer is >10,000

guggle• 6 years ago

I think Nim also has digit separator.

nimmer• 6 years ago

It does.

abrolhos• 6 years ago

C++ has an optional digits separator too. Not '_' but single quotes, since c++14 [0]. Which is cool. Swift has it too [1].

Erlang has the full stop '.' as expresssion/statement terminator [2].

[0] https://en.cppreference.com/w/cpp/language/integer_literal [1] https://docs.swift.org/swift-book/ReferenceManual/LexicalStr... [2] https://erlang.org/doc/getting_started/seq_prog.html

I like those features too! I'm glad they are coming to contemporary languages.

onychomys• 6 years ago

If we start using periods to end expressions, would we then have to use semicolons to chain things together? Or would you let the compiler know that foo.size(). was okay and that it shouldn't be waiting around for another function call there?

lifthrasiir• 6 years ago

Juxtaposition (`foo size()`) is a serious contender I think. In fact, I consider it is one of the most undervalued syntactic choices.

marcosdumay• 6 years ago

Hum... Prolog uses periods to end declarations, and as you expect, uses semicolons as combinators.

It reads really well, because it also uses colons as combinators, with higher precedence than the semicolon. But it does not map well into other paradigms.

pjmlp21• 6 years ago

Java has it since ages.

pansa22• 6 years ago

`_` digit separators are also supported by both Python and Go.

DFHippie• 6 years ago

To extend this list a bit, Rust and Ruby also allow _ as a separator.

macintux• 6 years ago

One is the use of the full stop as an expression terminator, as in English, instead of the semicolon.

That’s something people complain about when looking at Erlang code. Thanks to its Prolog heritage, it uses commas, semi-colons, and full stops to terminate statements.

It’s actually very easy to remember which is which, despite the general angst, because they are directly analogous to those punctuation marks’ usage in English.

Commas indicate a continuation of statements, semi-colons separate clauses within a function, and full stops complete a function definition.

mjcohen• 6 years ago

I also liked the "corresponding" modifier for verbs like "move" and "add". It caused the operation to only be done on those fields which existed in both the source and destination and ignored the others. Do any other languages have this?

ldeangelis• 6 years ago

Great article, I love learning more about language influence. I was wondering something:

Smalltalk wasn’t the only casualty of the “Javapocalypse”: Java also marginalized Eiffel, Ada95, and pretty much everything else in the OOP world. The interesting question isn’t “Why did Smalltalk die”, it’s “Why did C++ survive”. I think it’s because C++ had better C interop so was easier to extend into legacy systems.

Maybe it's linked to performance? Even today some application are rewritten from Java to C++ (or clones are made) to gain performance (like with Cassandra and Scylla).

linguae3• 6 years ago

I have another theory about why Java beat other object-oriented programming languages except C++: cost. I can't speak for Eiffel and Ada since I'm unfamiliar with those environments, but in the mid-1990s Java caused a revolution of sorts by providing free-as-in-beer runtimes and development tools that were available for download. I don't know how good GNU G++ was in 1995, but I know that Borland's Turbo C++ and Microsoft's Visual C++ were affordably priced. By comparison, commercial Smalltalk implementations had expensive licensing fees, and in 1995 Squeak and Pharo didn't exist. There was also OpenStep and Objective-C, but that was also very expensive; this 1996 article from CNet (https://www.cnet.com/news/next-gets-to-nt/) says that the Windows NT version of OpenStep cost $5,000 per developer and $25,000 for a version that allowed deployment.

With the high prices of Smalltalk and Objective-C environments, Java attracted a lot of companies and developers who wanted an object-oriented programming language that provided some of Smalltalk's benefits (e.g., garbage collection, memory safety, a rich standard library) without having to shell out the cash for a Smalltalk implementation.

pjmlp21• 6 years ago

I bought Turbo Pascal for Windows 1.5 (last release before Delphi) for about 150 euros, and Turbo C++ for Windows was about 200 euro, when converted for today's money.

This in 1990 - 1992.

jhbadger2• 6 years ago

But there was an affordable Smalltalk system in the early 1990s -- Digitalk's Smalltalk/V. It cost $99 and came with a huge manual that had a great tutorial. It introduced me (and lots of others) to the whole idea of object-orientation.

kqr5• 6 years ago

Ada was also very expensive up until relatively recently, so there could very well be something to what you're saying.

ansible• 6 years ago

Cost was likely an issue for Eiffel adoption in the 1990's.

There were a couple commercial implementations, with ISE being the dominant player. I bought a copy, but it wasn't cheap.

The open source SmallEiffel (later SmartEiffel) compiler by Dominique Colnet (and others) from loria.fr [1] didn't implement implement some parts of the language until the late 1990's.

It never had a big community.

[1] http://smarteiffel.loria.fr/

xkriva11• 6 years ago

Moreover, it is much safer to start to learn a new language/environment if you can immediately try it for free and not invest first into an expensive implementation.

igouy5• 6 years ago

…Java caused a revolution of sorts by providing free-as-in-beer…

Yes, when a well funded corporation gives away programming language runtimes and development tools — that puts language and development tool vendors out-of-business.

jerzyt• 6 years ago

In the mid 90s I wrote and supported a C++ program for moving around and processing data in a semiconductor factory. It ran on Windows and two different flavors of Unix. Half of my time was spent recompiling/relinking and trying to keep up with the upgrades of the few libraries I've used. When Java came out it was so liberating - I'd write a new version of the software on my laptop and ship an update to a client with HP Unix and it would run without a hitch. I doubt that there's still a lot of use of C++ for entire applications. I suspect it's mostly used in performance critical components.

pjmlp21• 6 years ago

Those rewrites don't get much uptake in many enterprises, most are more open to just rewrite a couple of modules into native libraries and call them from Java, than throwing away the whole eco-system and respective tooling.

Recurecur• 6 years ago

It is somewhat performance, but there are three other more important factors. (Java performance, not counting startup overhead, is fairly close to C++ now.)

One factor is determinism. GC introduces unpredictable (non-deterministic) time latency, making GC languages generally unsuitable for real time programming.

Another is size efficiency, Java programs have a well-deserved rep for using lots of memory. That isn't a good mix for smaller/embedded systems. Java also carries along a large runtime, although Graal and other efforts are addressing this.

The final factor is that C++ is perceived (rightly or wrongly) to be an "improved C". As the heir apparent to C, C++ has a ton of mindshare and momentum. Rust is now providing a major challenge, one I hope it wins! (Honorable mention to D, which is a nice language as well.)

linguae3• 6 years ago

Another highly influential programming language not listed in the article is Self (http://www.selflanguage.org). Although I've taken two graduate-level courses in programming language theory, I didn't learn about Self until just a few years ago when I was bitten by the Smalltalk and Lisp bugs and started reading about the history of these languages and their environments. One of Self's most significant contributions is the idea of prototype-based programming, which is the influence behind JavaScript's traditional lack of classes (though newer versions of JavaScript have support for classes).

I don't know if Self was ever commercialized; I do know that once Java was released Sun focused much of its attention on promoting Java at the exclusion of other object-oriented environments that Sun invested in (such as the OpenStep Objective-C API, which was actually jointly developed by NeXT and Sun). But Self is probably one of the most influential programming languages; it's just a shame that this language was never brought up in any of the computer science courses I've taken.

j-pb• 6 years ago

Self has an even bigger legacy, the JVM is basically all the technology that made Self fast ported over to Java. David Ungar pioneered not only Self and prototype based inheritance, but also generational garbage collection, and polymorphic inline caching.

Before Sun cannibalised the Self team into Java, it was actually slow as a snail.

Another fun fact, a reduced version of Self was used the programming language for most of the Apple Newton, because Dylan was taking too long to get off the Ground.

evgen4• 6 years ago

Having played with Self a bit before the Newton came out and then seeing it in NewtonScript, it was a very cool thing to watch the idea of system-wide prototype-based inheritance filter through the community. I think it was a bit too much for most people to grok, but the ability of a coder to modify only a few prototypes in the system soup to change the behavior of existing apps was absolutely amazing.

thequux• 6 years ago

Ungar's thesis is one of the most magnificent tours de force it's ever been my pleasure to read. It's mostly a series of "Here's why this technique is impossible to implement efficiently, but never mind that; here's how we implemented it efficiently anyway."

fanf22• 6 years ago

Then Lars Bak (who worked on Self and HotSpot) used the techniques to make JavaScript fast with v8

cwzwarich• 6 years ago

Self has an even bigger legacy, the JVM is basically all the technology that made Self fast ported over to Java.

Java only became fast after the HotSpot server compiler, which was written by Cliff Click (and Micahael Paleczny / Christopher Vick) and takes a different approach than Smalltalk / Self compilers.

SzamarCsacsi• 6 years ago

Classes in JavaScript are just syntactic sugar around prototypes.

BostonFern• 6 years ago

Ironically, what it really needs is explicit syntax for prototypes.

tannhaeuser• 6 years ago

This. And while we're at it, we shouldn't discount awk which had a profound impact on Self (and ultimately JavaScript) syntax, in turn. So much so, in fact, that the following pointless code (which doesn't even show JavaScript/awk regexp constants) is both JavaScript and awk:

    function f(x) {
      a[x] = "whatever"
      for (e in a)
        if (a[e] == x)
          return e
    }

Edit: but awk doesn't belong onto the list because it's far from dead

jonahx• 6 years ago

What were the specific impacts awk had on Self?

_asummers• 6 years ago

That's interesting that the curriculum didn't teach it! I went to a not terrific school and we learned about Self in undergrad P/L theory.

spdegabrielle• 6 years ago

The reputation of the school is a proxy. Teachers are the ones doing the teaching. I think you got lucky!

Scarbutt2• 6 years ago

I'm surprised Clojure isn't in the list either.

Edit: To clarify, Clojure is mostly a dead language that didn't have any innovations by itself, but it did influenced many programmers(the creator is good at marketing). It helped push forward the FP mindset into the users of other mainstream languages(js, python, java).

jacobush2• 6 years ago

Isn't Clojure a relative newcomer for it to influence language design in general?

A stepping stone to more mainstream languages perhaps? I hardly saw a Clojure program so please elaborate.

bonzini3• 6 years ago

Clojure is essentially a Lisp.

pansa22• 6 years ago

Is Clojure really “mostly dead”?

michaelmrose2• 6 years ago

This suggests we clarify what is meant for a language to be "mostly dead". If we consider size compared to total market or trajectory we may ultimately conclude that a project for example Clojure is "mostly dead" compared to JavaScript or java.

However this isn't ultimately the most useful metric. Would you for example select a restaurant based on the reviews and the cuisine or to total annual revenue of its parent company. If your metric suggests you forego surf and turf at a local diner in order to eat a Whopper you may be asking the wrong questions.

For your consideration here is a better one. A language is alive when its ecosystem is likely to receive enough interest and talent to continue to develop enough to allow its users to continue to accomplish useful goals. Being embedded in 2 of the top platforms and being able to reuse those libraries is a factor. Relying on these host platforms means that it can remain alive indefinitely while being useful to only thousands instead of millions as long as it drives enough interest to pay the salaries of the dozens who develop the core and the hundreds that write useful tools.

https://clojure.org/news/2020/02/20/state-of-clojure-2020

fulafel• 6 years ago

I think the mostly dead classification implies a decline, vs a non-mainstream mindshare of all developers.

ontologiae• 6 years ago

There's an interesting language strongly influenced by Self : Lisaac ( http://www.lisaac.org ). Lisaac is a purely prototype language extending the Self's prototype paradigm with extremely good performance (close to C code, biggest part of the research around this language was about high performance code generated by the compiler).

For instance, you can define a dynamic inheritance : your inheritance can be defined by a function which return an object - so, a prototype - among several.

The author of this language explain me that Self was one of the most important day in his life ;-)

diego_moita• 6 years ago

Lua's implementation of "object oriented programming" is also entirely based on Self's.

andybak2• 6 years ago

Lisp?

(ducks for cover)

I know someone writing Smalltalk on a daily basis for this system: https://kyma.symbolicsound.com/

diggan• 6 years ago

The article is about "mostly dead" languages. Lisp lives on under many forms, Common Lisp, Clojure and Racket being some of the more popular ones.

If we're only talking about influential languages, I'd agree with you and put lisp on the top.

andybak2• 6 years ago

I was being slightly facetious knowing the crowd round these parts.

agumonkey3• 6 years ago

and the never ending stream of emacslisp evaluations around the globe

QuesnayJr5• 6 years ago

That was an incredibly interesting article. I never quite grasped the historical significance of CLU.

Something I always wonder about COBOL, since it has so little influence on other languages, is if there are good ideas that we missed out on.

ngcc_hk2• 6 years ago

Guess online and efficient.

Done my 10k lines batch COBOL. It is a waste of life. Verbose. C is better but unfortunately it is not very io. Still ...

batch COBOL with mainframe parallel io sub-processor helps a lot.

May be the cics where 1000+ users can use a pc (3-5 mips even emulated) with 16MB is something unusual.

rebataur• 6 years ago

ABAP which runs in all of SAP systems in Fortune 500 companies as an ERP is derived from COBOL. The core product has 400Million lines. Extensions would be 4X

urxvtcd• 6 years ago

Interesting stuff!

Is there a kind of web archive of programming languages? Can we make sure their compilers, interpreters, etc. are not lost? Sure, there are issues of them being proprietary, or needing some virtualization to run; I'm just wondering.

pmcjones• 6 years ago

http://www.softwarepreservation.org/projects Is a start.

tyingq8• 6 years ago

Expecting to see Perl in an article like this in 5 to 10 years. Wasn't the first language I learned, but there's a fond place in my heart for it. At the time, it was the (only) less painful way to get at sockets, libc calls like getpwnam(), etc. I know the TIOBE index is flawed, but...ouch: https://www.tiobe.com/tiobe-index/perl/

PopeDotNinja• 6 years ago

And maybe Ruby in another 10 to 20?

goto115• 6 years ago

But how how influential was it? It seems most of the unique ideas in Perl have not been adopted by other languages.

But if Perl can be credited with kick starting the dynamic language boom (Python, Ruby) then it have been massively influential.

tyingq8• 6 years ago

I'd say it was pretty influential for Ruby, CGI-BIN, pcre, and other things that will live on in more refined forms. PHP has some obvious Perl influence as well.

Edit: Maybe also Perl's Configure (crazy wide cross-platform portability) and CPAN. They were pretty ahead of their time.

mattkrause• 6 years ago

I'd also give it some credit as a cautionary tale.

A lot of subsequent languages have focused on readability, partly as a reaction to some of the....unique things came up with in Perl.

xeeeeeeeeeeenu• 6 years ago

But how how influential was it?

Raku (formerly known as "Perl 6"), PHP, Ruby and PowerShell were directly influenced by Perl.

jeltz3• 6 years ago

CPAN is probably the biggest influence, it is the ancestor of most package managers used today. But Perl has also spread its dialect of regular expression and the idea of regular expression literals, plus it was hugely influential on Ruby which then have lived on in Elixir and Crystal.

Also you may be correct about it kicking off the scripting languages' popularity, and as someone else mentioned it is also responsible for cgibin.

mxyzpt1k• 6 years ago

Perl incorporated regular expressions into a general purpose programming language in a way that seemed fairly unique at the time. Today they seem to be everywhere, often with the help of PCRE-Perl compatible regexps.

downerending• 6 years ago

I think it would be fair to give Perl that credit. When it was first released, there really wasn't anything else like it. There was sh/csh, awk, sed, but nothing that really showed the path forward as Perl did.

That said, once Python 1.5.2 (or so) hit, Perl was done for. It was simply better, and there was no reasonable way for Perl to recover the lead, or even really coexist. Perl had a lot of momentum and took many years to decline, but that was the tipping point, in my mind.

It remains to be seen whether Python3 will displace Python2. :-)

timwaagh• 6 years ago

I have recently started to use it again as a tool for grepping across multiple lines. It's distributed with every Linux. So it will survive. Maybe it's already dead as a programming language though, certainly because Pascal is also considered dead.

pjc503• 6 years ago

Perl is possibly the clearest example of the "Osborne effect": announcing Perl6 took focus off Perl5, but it took far too long to arrive. So by the time it did many of the community had drifted away and the core use cases were becoming less relevant. Python almost did this to itself with the 2to3 transition.

cesarb• 6 years ago

An important difference is that Python 3.0 arrived soon, while it took ages for a complete Perl 6 release to be available. That is, if you didn't want to learn or use Python 2.x because it was soon going to be obsolete, you already had a working alternative; but if you didn't want to learn or use Perl 5.x because it was soon going to be obsolete, for a significant amount of time Perl 6 wasn't an option.

kqr5• 6 years ago

Perl consultants and distributors know that Perl is still widely used duct tape in the industry, and has even seen a slight upturn in popularity recently (although perhaps not as much as some other languages), when a new generation of developers has discovered how useful it can be. It's just not something one talks about.

TurboHaskal• 6 years ago

I like Perl a lot, it is the language I have used the most in my professional career. That being said, on my last two workplaces it happened to be in the list of forbidden technologies, not even being able to write a simple one liner and instead having to resort to using awk, sed and the like. It sadly suffers from underserved hate from developers who never used it.

amyjess4• 6 years ago

As someone who works at a company where Perl is used as duct tape, we've slowly but surely been ripping it out piece by piece and replacing it with Python (and by "we", I mean "me").

Much of our "duct tape" is unmaintainable code with undocumented aracne regexes everywhere, single-letter variables, and 1990s-era flow control. Nothing you'd find in Modern Perl is in our duct tape. Ultimately, the fastest turnaround on "we need feature X added to Perl script Y" is going to be "rewrite Y in Python, add feature X" because actually extending the Perl code has proven to be a nightmare.

(and as a side note, most of these Perl scripts ran out of cron on various boxen scattered across our network, and as I've been rewriting them I've also been removing them from cron and putting them in our centralized setup that uses Jenkins to run scripts, pulled out of git, inside Docker images on a schedule)

gerikson3• 6 years ago

Perl was probably culturally influential - scripting glue - but as a language it's almost uniquely eclectic. There are influences from all over, mostly shell and AWK but also from functional languages.

dleslie7• 6 years ago

The interesting question isn’t “Why did Smalltalk die”, it’s “Why did C++ survive”. I think it’s because C++ had better C interop so was easier to extend into legacy systems.

Video Games. From around 1999 to recently, C++ was the language of game development. It only recently came under serious threat, from C#.

frabert• 6 years ago

I don't really buy into the fact that enterprise apps are written in C++ because gamedev uses it

pjmlp21• 6 years ago

Very bad example.

Video games community is very luddite in what concerns adoption of technologies, they usually only move forward when the platform owners force them to do so.

Many moons ago, C, Pascal, Modula-2, Basic were seen as Unity is seen nowadays, naturally real game devs had to use Assembly, specially to take care of the special purpose graphic and sprite engines.

Playstation 1 was probably the first console that force them to start using C instead, while on 16 bit.

So slowly everyone accepted that doing games in C, Pascal and what have wasn't that bad.

C++ only became kind of accepted years later, and even then it was more "C compiled with C++" than anything else.

The major boost for C++ were the OS vendors for home and office computers, Apple, IBM and Microsoft, alongside Zortech, Borland and Watcom, with their full stack C++ frameworks, something that ironically C++ has lost (OS C++ SDKs where it has the front role).

davedx3• 6 years ago

I don't think it's a very bad example at all, it's just not the only reason C++ "didn't die out".

I worked as a C++ programmer from something like 2001 - 2010, and all video games companies used it. I wouldn't have learned how to program properly in it without the games experience I gained.

boomlinde2• 6 years ago

Video games community is very luddite in what concerns adoption of technologies, they usually only move forward when the platform owners force them to do so.

I think this is an unfair assessment. "Real devs" had to use assembly language because when targeting consoles and low-end home computers, this was the only really performant option for a long time. For a lot of types of games this doesn't matter so much, but there's a long ongoing trend in big titles really trying to cram as much visual fidelity as possible into the target machine.

Same deal with C++ today. You can use Unity of course, no one is any less of a "real game dev" because of it, but it's simply not an option if you want to work on cutting edge tech for AAA titles. There are tons of other things you may want to work on, other boundaries to push, which I think is why Unity is such a popular option. But I think it's unfair to simply chalk it up to luddism. Present the available alternatives that compare favorably in terms of development speed and performance for non-critical software instead.

dleslie7• 6 years ago

How is it a bad example of something which maintains the popularity of C++ when you acknowledge that C++ has been in use in the time period I mentioned, for that industry?

A few years after the Playstation gained prominence would be around 1999; and shortly thereafter Unreal Engine took the industry by storm.

EASTL is how old, now?

I still remember Carmack's comments of his initial forays into using C++ for idtech.

I don't think it's fair to shrug us off.

davedx3• 6 years ago

Also Visual C++ and Visual Studio. I worked on desktop software from around 2007-2010, not that long ago; all our software line was built in Visual C++ with MFC.

It's hard to believe how fast things changed and desktop apps were replaced by web apps...

pjmlp21• 6 years ago

Not everywhere, I am now back at doing Web related UIs, but have managed to keep doing desktop stuff every now and then.

There are domains, like laboratory automation devices in life sciences where browsers are persona non grata in air gaped environments, plus they lack the necessary hardware integration.

So between daemon + browser and a proper desktop application, most customers still rather have a nice desktop application.

FpUser4• 6 years ago

"It's hard to believe how fast things changed and desktop apps were replaced by web apps.."

Mostly business type front ends and it actually mostly makes sense for that purpose. But there are huge areas where browsers play no role.

ufmace• 6 years ago

I worked on Windows desktop apps around the same timeframe, maybe a bit later. IMO, C#/Winforms killed C++/MFC, due to being massively simpler and more productive, without giving up much in the way of performance.

skissane2• 6 years ago

Why did C++ survive the advent of Java? One reason: the mainstream JVM, HotSpot, is written in C++. As such, the production Java ecosystem is not self-hosting and relies on C++ to be sustained.

(Writing a JVM in just Java is doable, and some research JVMs have done it, but that approach hasn’t yet been fully productionised.)

pjmlp21• 6 years ago

GraalVM is taking care of that part, specially with Project Metropolis.

why_only_15• 6 years ago

There are probably >100,000 c++ devs out there, and ~100 of them work on HotSpot. I don't think the HotSpot devs are significant in encouraging c++ usage

Someone• 6 years ago

And that wasn’t because it had better C interop?

jeltz3• 6 years ago

No, while C interop was useful it was mostly because it was easier to write low latency code in C++.

ngcc_hk2• 6 years ago

Pascal have two moments which really peak. Turbo pascal and Apple Mac. After that not sure what its use. Ucsd pascal always slow.

Scarblac• 6 years ago

Delphi should be mentioned as well.

chadcmulligan• 6 years ago

Its not actually dead either

avian• 6 years ago

Does anyone know whether Turbo Pascal DOS/Windows compiler produced slower code than C compilers at the time? For a long time I was sticking with Pascal and resisted switching to C, even when everyone around me was doing so. One thing I do seem to remember from that time was that the things people were coding up in C seemed to work faster (not sure if it was Borland's Turbo C or something else). It was one of the arguments I remember for leaving Pascal.

Am I misremembering and it was just an urban legend/marketing from C compilers? Thinking back it doesn't seem logical that C would offer any significant performance benefits - both Pascal and C were compiled down to machine code and had approximately the same levels of abstraction.

pjmlp21• 6 years ago

Urban legend, as Borland fanboy, using their Pascal and C++ products, the generated binaries were pretty much the same, even if you might had to play with compiler pragmas, like disabling bounds checking for example.

In MS-DOS days, if you actually cared about performance, a macro Assembler was the only way to achieve it.

badsectoracula3• 6 years ago

While it is for Windows and not DOS, some time ago i wrote a small raytracing benchmark for C and Pascal (coded pretty much the same between the two languages) and the Borland compilers for C++ and Delphi (Object Pascal) had pretty much the same results: http://runtimeterror.com/tools/raybench/

What was probably the case is that there were more C compilers so some of them produced better code than Borland's.

fanf22• 6 years ago

The ML section misses out some significant successors: there was a fork between CAML and SML. Ocaml is still widely used, and has its own spawn such as ReasonML and BuckleScript.

LeonidasXIV2• 6 years ago

Neither ReasonML nor BuckleScript are really OCaml descendants. ReasonML is a different syntax for OCaml, with even closer match than CoffeeScript 1.x had to JS. Compiled ReasonML code cannot be distinguished from code that was written in OCaml. BuckleScript OTOH is a set of forks of older compiler versions which have a JS backend. It works with both Reason and OCaml syntax.

ReasonML therefore is much closer to the revised OCaml syntax (a failed previous alternate syntax) than it is to its own language.

tyingq8• 6 years ago

360 Assembler and Flash/ActionScript seem like good additions to the list.

muizelaar• 6 years ago

What did ActionScript influence?

greencore2• 6 years ago

TypeScript, I'd say.

tyingq8• 6 years ago

Arguably HTML5 and JS to get both in a state where Flash wasn't needed. Not just for what Flash is famous for either. I used it a lot for "copy this text in a div to the clipboard" when browsers were all over the place on support for that.

EamonnMR2• 6 years ago

If you're at all interested in this topic and have access to the ACM library, the 'History Of Programming Languages' (HOPL) conference includes papers on these and many other languages and is fascinating reading.

EamonnMR2• 6 years ago

And they just opened this up today! https://dl.acm.org/action/doSearch?AllField=HOPL&startPage=&...

ChrisMarshallNY• 6 years ago

I started off with BASIC, FORTRAN, and Pascal (as higher-level languages). I also used PERL and PL/1.

I don’t miss them too much, but they were important stepping stones.

gbrown_2• 6 years ago

At a time when adding two lists of numbers meant a map or a loop, APL introduced the idea of operating on the entire array at once.

Am I missing something, how is this different from a map?

7thaccount7• 6 years ago

APL by default operates on entire arrays even if they're multi-dimensional like a matrix. You can just say matrix x 3 and it will automatically do the nested for-loops behind the scenes.

gbrown_2• 6 years ago

Aha, thank you.

icen2• 6 years ago

It keeps going!

      3 3 3 ⍴ ⍳10

     1  2 3
     4  5 6
     7  8 9
       
    10  1 2
     3  4 5
     6  7 8

     9 10 1
     2  3 4
     5  6 7

      M ← 3 3 3 ⍴ ⍳10

      M*M

    1.00000E0  4.00000E0  2.70000E1
    2.56000E2  3.12500E3  4.66560E4
    8.23543E5  1.67772E7  3.87420E8

    1.00000E10 1.00000E0  4.00000E0
    2.70000E1  2.56000E2  3.12500E3
    4.66560E4  8.23543E5  1.67772E7

    3.87420E8  1.00000E10 1.00000E0
    4.00000E0  2.70000E1  2.56000E2
    3.12500E3  4.66560E4  8.23543E5

That's 3 nested 3x3 matrices, and we take the pointwise product of each of them with just the usual multiplication operator.

How many maps did you want?

RodgerTheGreat• 6 years ago

Incidentally this property is why most of the "translations" of APL expressions into another language you'll see online are cheating- they assume a specific rank for arguments and "bake in" an appropriate degree of mapping.

bonzini3• 6 years ago

Isn't that a pointwise exponentiation (1^1, 2^2, 3^3, ...)?

7thaccount7• 6 years ago

The APL section seems a little off. I downloaded the latest Dyalog version and the RIDE IDE and it was all very simple to use. Despite having zero APL experience I wrote a pretty nifty program to simulate something in my domain in like 3 lines of code. The keyboard thing is a non-issue as the IDE let's you enter symbols and you quickly start to memorize that you can enter command-key r to enter rho and so forth. I've used the ASCII J before and find the APL glyphs help me learn what's going on better. It's right that it isn't a popular language, but that is sad. Dyalog has a lot of tools too from web server, database libraries, R & Python Bridges, GUI...and so forth.

mlochbaum3• 6 years ago

I think it's fair to say that font issues were a significant reason for its decline in the late 80s and 90s (well before good unicode support). Other major factors were spreadsheets, which did many of the things APL was best at with an intuitive graphical interface, and OOP. APL didn't have OOP, so it was for dinosaurs. Structured programming could have had the same effect—mainstream APLs picked up ifs and while loops 10-20 years later than the rest of the world—but I think the usability gap between APL and anything else for arrays was just too large at the time for that to hurt APL too much.

mlochbaum3• 6 years ago

The article is wrong about mixing strings and numbers though. The requirement for homogeneous data, and the use of boxes, is a feature of the SHARP APL lineage, which includes J (although these languages could allow mixed arrays if they wanted to). But the APL2 and NARS families simply allow anything as an element of an array. These are much more popular, to the point that every recent APL I'm aware of uses this style of array. Possibly a reason why J wasn't very successful; probably not a reason why APL usage disappeared.

7thaccount7• 6 years ago

Thanks Marshall. I'll yield to you in anything APL as I'm a total noob.

For those that don't recognize his username, I think he works for Dyalog in APL implementation. Is that right?

WorldMaker8• 6 years ago

It's an interesting wonder if now that Unicode is much more ubiquitous, and people grow used to (and fond of) complex IMEs such as "emoji keyboards" and intricate ligatures of digraphs and trigraphs (fonts like Fira Code, Cascadia Code), if there will be an interesting APL resurgence or perhaps a modern Unicode "native" successor.

empath75• 6 years ago

I feel like APL would almost be better as a handwritten language.

7thaccount7• 6 years ago

So it was originally designed by a Harvard Math professor (Kenneth Iverson) as just that. A new mathematical notation to write down. He didn't get tenure (although later got the Turing award for some of that work) so he went to IBM and they actually implemented it. Aaron Hsu has done some blog posts on how he sometimes code using a nice calligraphic pen and paper with APL.

https://www.sacrideo.us/paper-is-dead-long-live-paper-progra...

amyjess4• 6 years ago

I'm surprised by how hard this article is on Algol 68. It was pretty influential:

- Influenced C's type system. I'm just going to quote Dennis Ritchie on this: "The scheme of type composition adopted by C owes considerable debt to Algol 68, although it did not, perhaps, emerge in a form that Algol's adherents would approve of. The central notion I captured from Algol was a type structure based on atomic types (including structures), composed into arrays, pointers (references), and functions (procedures). Algol 68's concept of unions and casts also had an influence that appeared later."

- Influenced bash's syntax (fi, esac)

- I'm not sure if this counts as an influence or not, but objections to Algol 68's design lead Niklaus Wirth to revive his earlier proposal for a new version of Algol, called Algol W, and ultimately evolve it into Pascal.

kick4• 6 years ago

That ALGOL-68 was hated by almost everyone who like previous ALGOL versions is not to be discounted.

ALGOL-68 was loved by people who would have never used ALGOL to begin with (and they didn't start doing so with ALGOL-68).

It was difficult to write compilers for, slow & far too complex.

ThomasBHickey3• 6 years ago

Looking back at ALGOL-68, it looks like comparatively small language compared to many of our current languages, e.g. Java, C++ and Python. I loved the definition of it, but never got to use it.

eeZah7Ux• 6 years ago

It's very similar to Go.

And like go, it's loved by people who don't know ALGOLs and hated otherwise.

yiyus2• 6 years ago

I'm sorry Forth is not on that list. It could be argued that it was not so influential for mainstream languages, but there is a whole range of concatenative languages that can be considered direct descendants.

carapace4• 6 years ago

I think Forth's influence is more subterranean. I.e. PostScript and coreboot, etc. Forth is there, but rarely in your face.

As far as the underlying concepts, speaking as someone who has experimented recently with Joy (one of the concatenative languages you mention. although it wasn't directly influenced by Forth, I think, it's a case of convergent design), I think it's a shame it hasn't been more influential.

The time may come: check out Conal Elliot's Compiling to Categories.

yiyus2• 6 years ago

Joy is a great language! I also wish it had been more influential.

Maybe you already knew this, but there are lots of great articles about Joy in nsl.com, and there is also Thun: http://joypy.osdn.io/ (which someone suggested me around here).

_0ffh• 6 years ago

Yeah, also the only other language that is as simple at it's core as Forth is Lisp. I suspect that lots of CS people have at some point written their own of either, or both, just for the experience.

gtk40• 6 years ago

I'm a younger millennial (born in early 90s) who got his start with BASIC. The first programming I did was on Q-BASIC on our aging desktop machine (386 with Windows 3.1) that we didn't use for much, especially with very limited Internet use.

When I took programming in high school, we started with TI-BASIC on TI-83 calculators for about a month, as my programming teacher felt like this best replicated his experience learning programming on a TRS-80. I tend to agree, and this is a great use of BASIC. It's the default programming interface on a widely used computer to this day.

We then moved to VB6 for our "serious" programming, although we also did JavaScript and Java.

My first programming professionally was done in an office setting at a temp job using VBA to help with some Excel work. And then my first job as a software engineer, even though I wasn't writing it, did have some Visual Basic.NET floating around (most of my work was in C#).

ORioN63• 6 years ago

I started to learn programming, by writing programs in TI-BASIC on the TI-84 (compatible with the 83) calculator. I just got it, because it was required to have a graphic calculator for math class.

It was just enough of a push. Small programs and games were fine and I actually got really used to the keyboard (I can still type on it pretty quickly these days). For longer programs I def. remember wanting to program with more monitor real estate and not having to rely on GOTOs. That's how I started to learn Python, which I still use daily.

nemo1618• 6 years ago

Same. My TI-BASIC magnum opus was 527 lines of spaghetti code, implementing dozens of nested menus for solving any trigonometry problem I was assigned. I honestly think that this is one of the best ways to introduce programming to a kid: "Hey, that math homework looks pretty tedious...wouldn't it be nice if your calculator could do all the work for you?"

gen220• 6 years ago

My story is similar to the OP's. Started with TI-BASIC, and downgraded to Java for a programming class. :)

It turns out, you can actually write programs in TI-BASIC on a computer, and sync the file to the calculator with the usb connector.

I used to charge kids in my class for copies of my games (Snake, and a Zork-esque text adventure). a lot of early business lessons there, in retrospect.

CarlRJ• 6 years ago

My first language was Basic (first on an HP-2000 timesharing system and then on the Apple II), but my second language was 6502 assembly, something for which I will be forever grateful - with the groundwork laid by assembly, C was a wonderful step up, handling all the tedious parts of writing assembly while allowing nearly all the same precision and control - reading the original K&R book was a transcendent experience. After those, other languages were easy.

dfan• 6 years ago

I took two software engineering courses at MIT around 1990 that used CLU as the implementation language (one was a general medium-scale software engineering course, one was on writing compilers). I found it to be a very pleasant language, although the main thing I had to compare it to was C (C++ was just getting off the ground). It's always nice to see it mentioned in lists like these.

thrownaway954• 6 years ago

Very surprised that ColdFusion isn't on this list. Say what you want about it, but remember this... You know how all you React, VueJS and frontend developers just LOOOOOVE components... Yeah... ColdFusion had that first in what was called "Custom Tags". Heck... It had alot of features that you find today in almost every other language, but it is dying a SLOOOOOOOOW death.

codingdave• 6 years ago

You are correct that ColdFusion pioneered some modern features, but the author had a specific definition of influence, and states: "Before we start, a quick primer on finding influence. Just knowing that X was the first language with feature Z doesn’t mean that X actually influenced Z.", and then goes on to state that influence only counts if it exists in citations of papers or documentation.

You can absolutely go to town arguing whether or not their definition of influence makes sense, though.

InfiniteRand• 6 years ago

It's questionable (although certainly possible) whether ColdFusion's custom tags inspired custom tags in other languages. There were a lot of other elements in the air at the time, such as server side include syntax, that made the idea of incorporating programming logic into markup a logic next step. And then from there it isn't too far to get to custom tags, of course, it's hard to retrospectively estimate the amount of brain effort to leap between concepts, so maybe my assessment is way off, but this is where finding direct documentary links between how technology developed can be helpful.

WorldMaker8• 6 years ago

ColdFusion had much more limited exposure to the general world than say PHP. Also, arguably and IMNSHO, ColdFusion was quite bad at a lot of its own features. Custom Tags specifically as the given example, were barely more than duct tape over SSI (Server-Side Includes) and never really had a "proper" Component model or DOM for most (if not all, again IMNSHO) of its history.

(CF is a language on my list of hopes to never see again.)

paulsutter• 6 years ago

I’m surprised that Ada isn’t listed

ThomasBHickey3• 6 years ago

Mentioned only in passing as one of the languages that Java sqashed.

azhenley• 6 years ago

CLU might be the most influential language that nobody’s ever heard of. Iterators? CLU. Abstract data types? CLU. Generics? CLU. Checked exceptions? CLU.

Well, I learned something new and I spend a lot of time reading about PL (and even teach undergrad PL!). Had never heard of CLU before.

restalis• 6 years ago

I'm still trying to reconcile "most influential" with "nobody's ever heard of". It may be that CLU implemented in premiere some concepts, but if later those concepts may as well just have been rediscovered (like the iterator design pattern) as the obvious thing to do, then what that leaves us?

carapace4• 6 years ago

reconcile "most influential" with "nobody's ever heard of"

We ignore the past. I've worked with people who hadn't heard of Alan Kay. I worked with a guy tasked to revamp an expert system who had never heard of Prolog.

WorldMaker8• 6 years ago

In this case there's a lot of evidence at least the language designers were paying attention to (Dr.) Barbara Liskov's published academic work over the decades, even if academic papers/languages rarely make it into mainstream consciousness. At a brief glance it looks like an Erdos number game of a sort, in that CLU papers were referenced a lot by other articles, then those articles were referenced by a lot more. There is probably a lot of one or two degree separation.

One specific dot this article helped me connect (thanks article) was that Barbara Liskov of CLU is the same namesake of Liskov's Substitution Principle, which definitely has made broader waves in mainstream OOP consciousness among programmers. It originated in a paper two decades after a lot of the CLU work so it isn't directly a part of CLU's own influence on the world, but it is quite probable that CLU's influence is deep inside the formation/elucidation of the Principle in that Barbara Liskov was clearly thinking about the problem area for decades, and in some cases decades ahead of colleagues and practical applications.

pjmlp21• 6 years ago

Generics, contrary to common wisdom it was CLU that actually introduced checked exceptions not Java, alongside Mesa one of the first modular languages.

celeritascelery• 6 years ago

I wonder which of these languages are completely inaccessible today? As in you can’t run them on any modern machine or even find an emulator for them.

goto115• 6 years ago

I would love some online tutorials/sandboxes for exploring these historical languages.

icen2• 6 years ago

It wouldn't surprise me if all of these languages were alive in some way or other. There are MLs and APLs and Pascals alive and well.

therealcamino• 6 years ago

That's an interesting question. I was going to guess CLU, but I was very wrong: there are native VAX and SPARC implementations, a Portable CLU implementation, and even a CLU-to-C compiler. Impressive for a language I hadn't heard of before today!

http://www.pmg.lcs.mit.edu/CLU.html

jbotz2• 6 years ago

Both ML and Smalltalk are definitely not dead, not even mostly dead, they’ve just evolved into language families whose members have different names. The ML family has Standard ML, OCaml, and Haskell, and the Smalltalk family has Pharo, GemStone, GNU Smalltalk, and others. These all may not have hugely wide adoption, but they are actively used in both academia and industry, and continue to grow and evolve, and their continued evolution is still influencing other languages.

Ok, you can argue about the definition of "mostly dead", but whatever you decide these two just aren't in the same category as the others on this list.

danharaj• 6 years ago

Haskell isn't in the ML family. It is a descendant of Miranda and some other lazy functional languages.

jbotz2• 6 years ago

Granted. 8-)

igouy5• 6 years ago

definition of "mostly dead"

It's easy enough to understand — just look at when they were very much alive.

"UbiquitousApplications:Embedded Systems to Mainframe"

http://www.davethomas.net/papers/ubiquitous1995.pdf

JoeAltmaier• 6 years ago

RPG? Or is that encompassed by Cobol...

padseeker• 6 years ago

No one is doing anything new in COBOL but there is so much old stuff that could only be replaced by a massive investment in rebuilding infrastructure. COBOL isn't dying anytime soon.

At one point, like 10-15 years ago, I knew experienced well paid COBOL programmers being laid off and being replaced by kids fresh out of school. And then CS programs, if not outright stopped, at least greatly reduced teaching COBOL courses. And no one coming out of school learning Java and Python and Node wants to write COBOL.

And then old timer COBOL programmers started retiring, but companies (especially banks) were not replacing their existing mainframe infrastructure. So now you have a gap between supply (COBOL programmers) and demand ( mostly banks), to the point that retired programmers are doing part time work for $200 an hour. Here's an article from 4 years ago;

https://www.reuters.com/article/us-usa-banks-cobol/banks-scr...

agumonkey3• 6 years ago

I wonder if IBM is still trying to migrate COBOL to something

kjs33• 6 years ago

I know at several dozen companies in my state alone who are still "doing something new" in COBOL, including the company I currently work for. No, that's not just maintenance, it's new apps and projects. Most of these are on a 'Fortune' list in terms of size.

Startups and small companies can ignore anything that isn't new and shiny; companies processing billions of dollars/transaction know better.

pjdemers• 6 years ago

Why I think Pascal is gone for commercial use. Back in the mid-80's I was working on a new project on PCs under DOS. Our first choice language was Pascal. However, sometimes we needed to write at low levels for performance, or to access hardware directly, or call Fortran code. For all of these reasons, we switched to C.

WorldMaker8• 6 years ago

At least one simplified view of the death of Pascal is the rise and fall of Borland. Borland's compilers were the fastest. Borland invested more in FFI work and C-interop than competitors. Eventually Pascal was Borland (and vice versa), because no commercial entity (in the 90s-ish) was using a non-Borland Pascal compiler. Borland making bad business decisions and Microsoft making good business decisions in poaching key Borland staff at the right moments, seems like some obvious death nails in Pascal's coffin. (Obviously that's still a very simplified narrative to history, but it is an easy and common one.)

dboreham• 6 years ago

Anyone else here written BCPL? I must be one of very few people to have used it at _two_ different jobs.

dnautics2• 6 years ago

I love both languages (well c99 anyways), but I feel like Pascal/C is like betamax/vhs.

I think I it didn't help that free Pascal compliers were harder to come by. IIIRC

throwaway17_17• 6 years ago

I’ve seen multiple statements from people that Pascal was a better language than C, but I absolutely love the VHS/BETA-Max analogy (it contains a lot of context which usually take a good deal longer to write out in the pascal vs c discussions). My education and experience effectively skipped Pascal, other than mentions that it existed. So, do you have any off the cuff reading or materials that demonstrate where Pascal shines? I’d be interesting to see if any of Pascal’s qualities that made it ‘better than C’ were ripe for plucking out and putting in another language.

dnautics2• 6 years ago

I haven't used Pascal in 20 years. I learned it in cs/AP cs (literally the last year it was on the apcs exam), used it for my high school science fair project and in my senior year I self taught and rewrote it to c++ and never went back.

Looked at Eiffel for a hot sec in college but it never stuck, because no toolchain for beos, which I used in college.

These days I program in elixir, and am learning Zig.

TomMasz• 6 years ago

I did my Master's (1994) work with Smalltalk (specifically Little Smalltalk) and I still play around with Squeak when I have some free time. Not exactly producing production code, obviously, but you can build things for yourself and there's nothing wrong with that.

mkchoi2122• 6 years ago

"almost all of logic programming actually stems from Prolog"

Why isn't Prolog on the list then?? :O

detaro2• 6 years ago

is Prolog deader than the logic programming systems stemming from it? The field seems fairly niche overall, but from the limited insight I have Prolog seems to be still around?

mkchoi2122• 6 years ago

My thought was that prolog is now mostly used in colleges around the world to teach students basic concepts within AI. Correct me if I'm wrong though. Would love to see examples where Prolog is used in real life!

btilly• 6 years ago

I am sad that Forth did not make the list.

It showed how to make a minimal programming language that runs in a very small amount of space with just a stack. And did a lot to popularize RPN notation.

Even if you do not write in Forth, you can still benefit from knowing the ideas. For example I could not have written my answer at https://stackoverflow.com/a/60817908/585411 if I did not know the ideas of Forth.

vanderZwan• 6 years ago

It was probably considered too "alive" in the embedded world to be on the list.

lucas_membrane• 6 years ago

Nice article. Missed one thing about PL1 that I think it got first and that you see nowadays in most languages (other than low-level languages like go and C) and most large programs -- a hierarchy of error types, blocks of code to catch errors (exceptions) according to type, and the ability to replace one error-handler with another temporarily within a narrower scope. Not to say that this is good, but it is what we've got.

V-2• 6 years ago

"COBOL was also enormously complex, even for today’s languages"

Is that true? What was the complexity about? I've never done COBOL, but I was always under impression that it was quite straightforward in its infamous clunkiness, lacking lots of constructs that've been obvious for decades now

agumonkey3• 6 years ago

I don't know but there's the ALTER (and maybe a COMEFROM) ?

peter303• 6 years ago

Reads like the graveyard of my high school and college languages

BASIC 1970 high school computer hacking

LISP 1972 MIT A.I. course

APL 1973 MIT circuits course

PL/I 1973 MIT systems programming course

OS360 Assembly MIT systems programming course

FORTRAN 1974 MIT physics research

C 1976 Stanford research

PASCAL, ObjectPASCAL 1984 Macintosh developer

ObjectiveC 1987 NeXT Developer

C++ 1986 scientific programming

Java 1994 scientific programming

rst• 6 years ago

For all the mentions of FORTRAN in the description of other languages, it's a real surprise that FORTRAN doesn't rate a mention of its own -- though there are corners of the world where it's still in use.

It's notable not only for being the Lingua Franca of scientific computing through at least the '80s, but also for breaking ground in programming language technology, starting right at the beginning -- the Fortran I compiler, for the IBM 704 in 1957, was the first to have an optimizer (and the technical papers on that compiler introduced terminology which has since become standard to the field -- e.g. "basic block").

kjs33• 6 years ago

If you leave the HN/SV bubble, you find huge amounts of Fortran. It's still plays a significant role in scientific computing. If you lost track of it in the '80s, you might be interested that the language had major updates in '90, '95, '03, '08 and 2018. It's just not a part of the world of cat pix sharing websites and innumerable "Foo of Bar" startups.

Same with COBOL.

couchridr• 6 years ago

One thing I credit COBOL with is the use of descriptive variable naming. In FORTRAN you see variables named i, j, x etc (just look at Numerical Recipes) but in COBOL you see TOTAL_MONTH_SALES and the like. It is considered good style and “self-documenting” to do that today.

hiddenhandle• 6 years ago

No mention of Snobol? Precursor. Only saw it in texts though of Perl

mbo• 6 years ago

This article echos a lot of the themes that I saw in "Family spaghetti of programming languages" (https://erkin.party/blog/190208/spaghetti/) which seems to be the best attempt at developing a family tree of programming languages I've seen thus far.

amyjess4• 6 years ago

Ooh, I really like the idea of this, but I kinda wish he broke Fortran into a whole family, with separate entries for each major version. There was a lot of cross-influence between Algol and Fortran, and while the original Fortran inspired Algol, Algol ended up as a massive influence on how Fortran developed later.

Most(ly dead) Influential Programming Languages

Related Stories