[phicoh]

The article gives another reason "A second answer is aesthetic. Ada's syntax is verbose in a way that programmers with a background in C find unpleasant. if X then Y; end if; instead of if (x) { y; }. procedure Sort (A : in out Array_Type) instead of void sort(int* a)."

I think this should not be underestimated. There is a huge number of small C compilers. People write their own C compiler because they want to have one.

That doesn't happen we Ada. Very few people liked Ada enough that they would write a compiler for a subset of the language. For example, an Ada subset similar to the feature set of Modula-2 should be quite doable with a modest effort.

[lproven]

> I think this should not be underestimated.

You're right but it's broader than "C folks like terseness."

C is famously hard to read. Before Perl we used to joke that C is a write-only language: you can't understand what your own code means just weeks later.

Combine this with its lack of bounds checking, pointer arithmetic, and other dangerous features, and the result is a language that's macho for geeks: it's hard, it's dangerous, but it's small and it's fast.

It's a motorcycle for nerds. Ada is a tank.

Nerds get to establish dominance over lesser nerds by doing hard stuff in hard languages and making it fast. This bestows nerd street cred: geek cred.

Ada was used by contractors who needed stuff to work and money was no object.

C was used by hackers to do cool hacker stuff that was perceived to be fast and low level.

It's not low level: machine architectures haven't resembled the C abstractions since the 1970s.

https://queue.acm.org/detail.cfm?id=3212479

A modern low-level language would be some brain-bending combination of APL and Lisp with n-dimensional tensor algebra or something.

But C looks cool and hard and you will blow both feet off if you don't hold it just right.

And there are good free versions. So you can be poor and still demonstrate your machismo.

Result, a software industry requiring weekly multi-gigabyte online patches, keeping millions in work.

C makes programmers a cheap fungible commodity.

https://www.loper-os.org/?p=69

[bigbuppo]

The real problem is that Ada forces you to plan ahead and most developers don't really know how to do that.

[phicoh]

I'd say that is even more so with Rust and Rust got popular in a very short amount of time.

[brabel]

I think this was a genuine generational change. I am pretty sure Rust would never have become popular 20 years earlier because the priorities back then were so different (that was the era of languages like Ruby and Pearl where conciseness and low verbosity were the most valued aspects).

[jordanb]

When Ada came out a lot of programmers couldn't even touch type. You're right there's a generational change and a lot of of the Ada stuff won:

    * strong typing
    * lots of annotations
    * keywords over syntax, support for long variable and token names
    * object focus (Ada 83 had some limitations on inheritance so it wasn't OO strictly speaking) 
    * exceptions
    * large standard library

These things were controversial in the 1980s. They are not today.

[phicoh]

I think that is not correct.

One of the big differences between K&R C and C89 is the introduction of function prototypes. Strong typing was certainly considered positive for compiled languages. Of course C is a lot less strict than Ada.

If we compare the Rust subset that has similar functionality as C then there is not much difference. You get 'fn'. The is 'let' but Rust often leaves out the type, so 'int x = 42;' becomes 'let x = 42;' in Rust. Rust has 'mut' but C has 'const'. Rust introduced '=>' and removed '->' from object access and moved it to the return type of a function.

The C language has support for long variable names. Some early linkers didn't, but that's an implementation issue, people were certainly unhappy about that.

C++ started in the 80s. Objects were not controversial back then. The same applies to exceptions.

I don't have a metric for the size of a standard library. For its time, the C library in Unix system had a large number of functions. Later that was split in a C standard part and a POSIX part. But that was for practical reasons. Lot's of non-Unix systems have trouble implementing fork().

I have no clue what you mean with annotations. If you mean non-function annotations along with code, then generally Rust programs don't have those.

[jordanb]

Exceptions were controversial into the 90s which is why Java went down that whole checked-exceptions rabbit hole. The argument was that an exception was essentially a GOTO (or even COME FROM) which broke functional abstraction.

The Ariane 5 crash involved an exception and that was the central "Ada is unsafe actually" argument from C people.

In fact "exceptions are bad" is so baked into a lot of C people's brains that they left them out of Go!

Short variable names were a technical limitation in early languages but style guides were still arguing against long, descriptive variable names in languages like C into the 2000s.

Objects were also likewise controversial and you can see that in the design of Ada 83 where they were both inspired by OO languages like smalltalk but also hesitant to adopt stuff like inheritance. Inheritance was again, seen as a way to break encapsulation (it kinda is) but also a lot of object implementations were slow and memory inefficient in the 80s. Smalltalk was pretty much the reason why the Apple Lisa failed as a product.

OO became a massive buzzword in the 90s but by that time it had already been around for quite a long time.

By annotations I mean mostly type annotations, of course there's also aspect annotations and other stuff ex: Ada SPARK.

[pjmlp]

Function prototypes were actually taken from the C++ ISO process, back into C, originally.

Back when I learnt C, I think you could go beyond 8 or 12 characters in the symbol tables of compilers like Small-C.

[pjmlp]

As Gen-X, in the Usenet flamewars, the C and C++ folks used to call Pascal/Modula-2/Ada advocates as straightjacket programming, whereas they would be called cowboy programmers.

Ironically the author of Fil-C calls classical C, YOLO-C. :)

[budman1]

not just the priorities, the overall skill and education of programmers.

in the 1980/1990's i was a dumb kid. problems of large systems were not in my mind. having to type begin/end instead of {} was, i thought, a valid complaint.

with experience, education, and hindsight, most of the advantages of the ada language were not understood by the masses. if ada came out today, it would have taken off just like rust.

[phicoh]

I'd say that if the original Ada was introduced at the same time as Rust development started then people would pick Rust. Ada is also a product of its time would have to be modernized quite a bit.

Given how similar the syntax is of C, C++, Javascript, and Go, I think a language with the syntax of Ada would have a hard time.