Slashdot Mirror
Why Apple and Google Made Their Own Programming Languages
Gamoid writes: This Business Insider article looks into the state of Google Go and Apple Swift, highlighting what the two languages have in common — and why tech companies would bother involving themselves in the programming language holy wars. From the article: "One fringe benefit for Google and Apple is that making your own programming language makes recruitment easier — for instance, since it builds a lot of its own server applications in Go, Google is more likely to hire a developer who's already proficient in the language since she would need less training."
That sentence is actually valid in Brainfuck language.
839*929
Really. That last sentence proves it. They have no fucking idea what different languages are good for, or not.
It weeds out the people who aren't fan-boy enough to become proficient in your proprietary language before you even interview them. TFA cites a lot of other reasons they wrote these languages. This one got crab-apple picked for some reason.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
From wikipedia
Go's "gc" compiler targets the Linux, Mac OS X, FreeBSD, NetBSD, OpenBSD, Plan 9, and Microsoft Windows operating systems and the i386, amd64, ARM and IBM POWER processor architectures/ A second compiler, gccgo, is a GCC frontend.
So there are two major compilers for Go already, one of which is gcc based which targets just about every platform under the sun. I'm not saying go will run everywhere gcc will compile code because the runtime also needs porting, but it is very cross platform.
I developed one of my command line apps in Go http://rclone.org/ and I release binaries for it which run on Windows, OS X, Linux, *BSD and even Plan 9 all cross compiled from my Linux workstation.
Every man for himself, all in favour say "I"
This is a bad article; and the submitter, editor, and readers should all feel bad.
The reason why companies develop new languages is because the ones coming from academia are focused on the wrong things. Product development requires an industrial strength, strongly typed (for the most part) fast language.
Projects coming from academia are interpreted, JVM based, functional, obsessed with (im)mutability, closures, and lambda functions.
This is not to say those things are not nice, however they are not central to a programming language as used in large scale industrial systems.
Its not even logical. If the goal is to hire developers who need the least language training, having your own in-house language is the worst way to go. Clearly google and apple should pick the most popular language if that's what they want.
On the other hand, an in house language probably reduces developer mobility -- they can't go to work for another company if they don't work with common languages.
Add to that, Go and Swift are pretty small languages. Learning either is something that a moderately competent programmer ought to be able to do in a few weeks. Neither is sufficiently different to other languages that there's a big cognitive jump. The difficult thing is always learning new libraries and frameworks, not learning a new language (well, unless the new language is C++, where after a decade of daily use developers are still not surprised to come across a language feature that they've never seen before).
I am TheRaven on Soylent News
is it really impossible for anyone to believe that a language and toolchain can actually
make an organization more productive?
it seems like everyone is so lost in technical marketing that they've forgotten
about actually programming computers
Exactly, this doesn't make sense at all for recruiting, it's actually backwards: using your own programming language makes it far more difficult to recruit, because very few people from outside your organization will have any expertise in the language.
It doesn't matter how many platforms Go compiles for; what matters is how much demand there is for Go programmers. Outside of Google, the answer is approximately zero, so it's a bad choice if you're worried about your continuing employability. C++, on the other hand, is used all over the place, so it's a very safe bet.
Anyone who is proficient in Swift has hundreds or thousands of companies outside of Apple looking to hire them. Apple has to compete with every iOS software development shop in the world for those people.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
I hope they guan get visual studio working on other platforms
To that end, you could try out Visual Studio Code, which was introduced at the build conference this year.
"They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety."
The makers of this article clearly have no background in computing, or journalism either for that matter. I'm surprised I didn't see a reference to the Illuminati in there somewhere. Bizarrely, the article doesn't even mention Dart, which is no doubt due to the two-minute Bing search that I'd imagine formed the entirety of their background research.
When considered against the status quo for their purposes and eras, all of these languages show significant, useful advances in programming. And if we're going to declare all languages that are created by a for-profit corporation invalid, say goodbye to Java, C#, C++, and C. Hell, even the Jacquard Loom was meant to make money.
It took me less than a week to learn go. I still don't know all of C++ after two years. Learning Go was not a bad investment: the perspective I learned from it helped me write better C++ and C# code from both a OOP design perspective and threading perspectives, and is also a nice thing for my resume (Maybe is helped me get my option to work for Google, I don't know).
If you've made your own language, you're more likely to keep your experienced employees because there are fewer places for them to go to if they jump ship. Sure, there will be companies that use the language because they have to develop software for Google or Apple, but the employees are going to be a lot more locked-in than if they were experienced in something more widely used such as C or Python.
There is a minor recruiting benefit, only to "go".
WHY SWIFT IS NOT BENEFICIAL:
Internally, Apple doesn't write a lot of Apps; they might, eventually, convert some of their existing large applications, like Pages or Mail.App, over to Swift; these will largely end up being rewrites. So while knowing Swift might make you better able to get a job at a software house targeting Apple's platforms, it's unlikely to be meaningful in getting a job at Apple.
On the other hand, one benefit to Apple is that if Swift is not strongly used internally, the demand for App developers is unlikely to hire away Apple engineers out from under Apple, which was a big issue with the Objective C "brain drain" when iPhone and other apps took off. One of the things that Apple did, for example, was not let registered developers who were also Apple employees, take App programming classes held by Apple, for a period of six months after they started offering them to non-Apple employees. So there is an "anti-recruiting away of Apple employees" benefit to Apple.
WHY GO IS BENEFICIAL:
The recruiting benefit of "go" is clearer, although even with gccgo, go is not very portable to non-Linux platforms, despite its claimed platform support (for example, the standard libraries *still* have some serious compatibility issues on Mac OS X, despite the fact that almost every Google employee has an Apple laptop).
One thing that companies like to hire is young people; on the theory that "you can't teach an old dog new tricks", someone who knows "go" is a lot more likely to be younger, rather than older. It's not valid due to the theory, but it is nevertheless valid enough that you can pretty much use it as something of an age filter, and legally get away with doing do.
Another thing it means is that you're willing to learn new things; a surprising number of people aren't. If you got into software engineering for the money, and you are just using it as a "paycheck continuation program", instead of actually being passionate about it, it's unlikely that you've bothered to take the time to learn "go".
These are relatively minor benefits, since it doesn't take that long to learn a new language well enough to work in it pretty extensively; so training is not really a benefit, as the article claims.
But they don't have a shortage of applicants. They actually want to narrow the field. If that were not the case, Google wouldn't have had (perhaps they still do) such a notoriously difficult interview process.
Economies of scale are critical here. Only a handful of companies are that big, and that desirable as places to work. So for these behemoths the usual logic is inverted. For them, narrowing the field really does "help recruitment"--the semantics of that phrase are inverted when dealing with relativistic money.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
OTOH, it's easier to retain employees who only know "your" language.
One fringe benefit for Google and Apple is that making your own programming language makes recruitment easier --- for instance, since it builds a lot of its own server applications in Go, Google is more likely to hire a developer who's already proficient in the language since she would need less training.
And had Google used C, it would be more likely to hire someone who's good with C, since they would need less training.
Of course, C# has a much larger ecosystem, which is also kind of the point.
Of course, C# is also Microsoft's baby, which is enough to disqualify it from use inside Apple and Google all by itself. Nobody at either of those companies is going to risk relying on Microsoft's intellectual property for anything mission-critical.
I don't care if it's 90,000 hectares. That lake was not my doing.
Because if they used, for example, Java, when they hired an experienced Java developer they would need MORE training?
Approximately how large is the pool of 'experienced' Go/Swift programmers outside of Google/Apple?
Ken
"I'm still scratching my head over the use of "she" instead of "they" in that sentence."
It's Corpspeak. You have to use a precisely equal number of male and female pronouns, even when that means shifting gender confusingly back and forth throughout your Powerpoint presentation. But if your presentation contains a example of bad procedure, you have to use a male pronoun at that point ("If the nurse were to stick his finger in the 220V socket during Step 5...")
He means "I tinkered with this new thing and it seemed pretty easy enough, so that was enough - I never needed to use it, so I never found out all the edge cases I know about in the mainstream languages I use every day".
Its why many people like new languages and things, they think they're simple because they only have to scratch the surface for some simple example and think that's all there is to it.
I'd define 'learn' as be able to read 90% of the code that's out there. And yes, I can easily see C++ taking years by that standard.
"Yes, our code *does* overload the 'space' character, I thought you said you knew C++!" :-).
Welcome! We're always glad to see representatives of HR visit us on Slashdot. Next week as you write a requirement for five years of Swift into that new job posting, you can thank us for having inspired you in your quest for offense-free skill categories.
Masculine pronouns can be used in the gender neutral sense in English. Feminine pronouns cannot. Posting as AC because I don't want saying obvious fucking truths archived on my slashdot account, thanks to the witch hunts political shitfucks will inevitably engage in.
He means "I tinkered with this new thing and it seemed pretty easy enough, so that was enough - I never needed to use it, so I never found out all the edge cases I know about in the mainstream languages I use every day".
No, that's not what I meant. After spending a week learning Go, I wrote a couple of applications in it, but after the first week I was not learning Go, I was learning libraries and patterns.
Read the Go spec: these is almost nothing to it. The language has ridiculously few features. The main design goal of the language was to make code very readable and maintainable which means all devs basically know the entire language spec.
Don't read the C++ spec. You don't have time for that and it keeps growing. C++ has lots of features, which is both a good and a bad thing, but not all languages are like that.
Plenty. All of your criteria have little to do with the language and much more to do with the developer. Any properly defined/debugged program is "safe". Any properly optimized program is "performant". "Clear" is just about source code, which means that it's entirely up to the developer.
ISO 9899:2011.
ISO 14882:2014E.
ECMA 334.
Swift is a "lookalike" to all of these in several ways, especially as that list goes on. The list of languages that aren't ancestral to Swift but that have standards could go on for quite a while longer.
Very well, actually. ADA is very good at what it does well, which is creating deterministic code that can be procedurally validated. It was created for a design environment that is the exact, deliberate opposite of Agile. If you think you should build once a day or week, then ADA is not for you. If you can not simulate your live environment because it's a thousand miles up in orbit and a bug costs half a billion dollars and 2 years to the next launch window, ADA is for you. If virtually memory machines are a pointless risk and dynamic memory allocation is anthema, then go for ADA. If not, stay the fuck away, because it will make you a grey beard, one way or another.
It reduces the pool of resumes that you have to sort through into just previous employees and liars. Since you know that the previous employees are either no good or have a grudge against the company, you can filter them out and hire the rest for sales.
Economies of scale are critical here. Only a handful of companies are that big, and that desirable as places to work. So for these behemoths the usual logic is inverted. For them, narrowing the field really does "help recruitment"--the semantics of that phrase are inverted when dealing with relativistic money.
A filter is only useful though if it removes the bad applicants and leaves the good applicants. Filtering by language (/framework) although common is also a very good way to exclude a significant amount of programming talent on the basis that you don't want to give them a few weeks to get productive in your pet language/framework. I've never interviewed (or applied) at either but both Google and Apple seem to have more farsighted hiring practices than that.
*fewer
...and it had nothing at all to do with recruitment.
When Apple came up with the LLVM static analyzer, they became much more aware of just what kind of coding mistakes were costing them the most time and money, and Swift addresses those issues.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Yes, it's very unfair to men that they don't have their own pronouns.
Have you not noticed that almost all programmers are men? That means that when you use the word "she", you're saying it only to make a political statement, not because you think it actually describes the situation.
Disclaimer: IANAL. This post is, however, legal advice, and creates an attorney-client relationship.
If you won't acknowledge that the language can guide users to make the correct choices then I think you're missing the point. By your definition ASM is safe and clear. Personally I prefer to have projects coded in a language that encourages expressiveness and convention, but that's mainly because most of my time is spent on systems that are designed to live for 15-20 years. Reducing or eliminating maintenance time is more valuable to me than pure wire-level performance. Others may have different trade-offs. You should know your project's goals and choose accordingly.
You're special forces then? That's great! I just love your olympics!