This relatively recent trend of company-specific languages annoys and disturbs me.
I don’t ever want to be tied to a language and library ecosystem under the thumb of a single (large) corporation. Not Visual Basic, not .NET, not C#, not Objective C, not Go (It’s even named after the company, for crying out loud! Yes it is. Don’t try to claim otherwise).
I’ve used Basic, I’ve used Pascal, I’ve used C, I’ve used Python, I’ve used Lisp, and so on and so on. Those were open platforms, with different companies in the “lead” position in any one time (Microsoft Basic, Turbo Pascal, Borland C, GNU C Compiler, CPython, PyPy, etc.), but the language was not “owned” by a single company which would loom in everyone’s mind, always the unspoken fear being “what if <company> doesn’t like it?”.
I will not use tools which make me afraid. I will not live a life in fear.
I'm generally in agreement, but it's hard to say what languages will be considered "company-specific" in the future.
Go back in time 35 years and C could have been called that "AT&T specific language" Or the opposite: Objective C looked like a hopeful contender to be a widely-used language but it just never caught much traction outside the NeXT/Apple (and NeXT ended up buying the developer) and today it's almost completely associated with a single company's ecosystem.
Maybe in 4 years we'll be talking about the new features coming in "ISO Go19" and complaining that VisualGo still doesn't support all of Go18 yet. Who knows?
True, but today Go is still specific to Google. If it really catches on in the way you describe, then it will be a viable language. But not today.
But look at other examples.
JavaScript became ECMAScript, where ECMA is there to define standards, but still there are a couple of big players from certain companies and still they do Dart, which is mostly a Google thing and is a direct competitor to ECMAScript.. well, ECMAScript 6 at least. Even though it's not a bad language one can see a lot of company interest there.
And when you look at Go you see that it is controlled by people who work for Google, but don't exactly depend on them. I think it's more a language of Ken Thompson or Rob Pike than Google.
What do you think about Rust? It's affiliated with Mozilla, but Mozilla is (also) a Foundation and Rust is not a Mozilla Product like Firefox, where it often is really hard to get your changes in. There are a lot of people who got interested, even outside of Mozilla and at least to me it appears that they have influence.
And the last thing is what matters. Microsoft has huge interest on controlling C#, Visual Basic .Net, etc., because their income is related to data. They create the major platform for it, get income via selling the OS, the IDEs, etc.
If you look at Go then that doesn't seem to be the case. Google's main benefit (tell me if I am wrong!) appears to be being able to replace C++ and Python where neither seems to fit for one reason or another. If you look at the state of both the Android version now and Google App Engine then Google doesn't show huge interest into using it for money.
Of course that is only now and maybe this changes, but it doesn't look like it for now.
My personal opinion is that Dart was more to worry, also because of the approach being similar to Microsoft, when they create a standard (be it for documents or a language) and then create their incompatible extension or are always ahead of the others, because of their strong involvement. I really don't see that with Go.
Another thing that is slightly related to go is NaCl (Native Client, not the crypto library), being Google's version of ActiveX. But that's more because it of course is a target platform for Go, not a thing that's wrong with the language. Else you'd also have to worry about C.
Since you ask: Rust, being a Mozilla thing and Mozilla being a foundation, is certainly a less risky and “fearful” (so to speak) platform in this regard. Go, however light a hand Google might have on it at the moment, still has Google’s hand on it, which makes me wary. Sure, it’s better than Dart (or C#), but everything can point to something which is worse – that is not a good enough reason to accept a thing’s drawbacks.
If the non-company-specfic languages cannot keep up with the evolution company-specific languages show they will have a problem getting or keeping their users. I know it takes a lot of investment to build great tools, do marketing, evolve a language etc and currently some large companies seems to think that they get more of what they want by funding their own languages. If someone fund the free languages they same way they will probably evolve faster and get more users. There is also a problem of getting large groups of people and organizations to agree on something. This also slows down progress in some open/free projects. Sometimes it is easier to just decide for yourself.
There are certainly advantages for companies to develop languages in-house, but there are also, as I wrote, disadvantages for outside developers to use that company-specific language. This lesson was learned long ago about assembly languages and machine code: If you could program the Burroughs Whatzit 220, you could not program the Data General WhizJig 3000, and all the software ecosystem you (and others) had written over the years became obsolete when technology advanced. This was a large part of why programming languages were invented, and why they were always meant to be platform and vendor independent – so that this would not happen again¹.
It’s possible that we are due for another generation of developers to make the same journey of discovery of why vendor-specific technology solutions have larger drawbacks than what is initially apparent.
① This is also, incidentally, why Unix proliferated as an operating system – it was and is, for all intents and purposes, a vendor-independent platform for development at the operating system level.
>but the language was not “owned” by a single company which would loom in everyone’s mind, always the unspoken fear being “what if <company> doesn’t like it?”. >I will not use tools which make me afraid. I will not live a life in fear.
This is a very emotional response. It uses the language of emotion: fear, afraid, even the odd emphasis of the words owned and loom. Unfortunately for you Hacker News tends to be logical and analytical.
Personally I don't fear any tools, except for table saws, which combine extreme sharpness with a dangerous axis of rotation that can lop off fingers or propel a 2x4 through your abdomen with ease.
Oh, I also fear C++, which is the table saw of the mind.
It's open source.
While technically true, this has almost nothing to do with what I wrote. Huge code bases thrown over the wall, while technically “open source” does not a community make. A language named after a company with the overwhelming majority of development sponsored by that same company is not something you realistically envision someone forking, so the fear I spoke of is still there.
Now, if the language was, say, maintained by 80% or more by submissions and monetary contributions from outside the company (and if the name was changed), then it would approach being a neutral platform to the benefit of all.
As it is, we’re all being invited into Google’s yard to play, but we don’t own the house.
How would communities around anything come to exist if everyone had the same attitude?
Look at how the communities around languages like Basic, Pascal, C, Lisp and Python have developed.
They were, from the beginning, open with development, releasing early and often, and were explicit about readily accepting large contributions from outsiders (and really did so). The creators were always open to the possibility that they themselves might not be the eternal keepers of the language, which allowed competition when others developed the language further.
Contrast this with the development of the languages I mentioned. They have done the opposite of these things.
I think you should do what you ask to be done. You fork Go under a new name, create a community with the same objectives as yours around it, be competitive and develop the language further and then reap the satisfaction. If there is a significant need for your ideas I'm sure the developers will join you (otherwise you will have the best proof that your idea is just not good enough for many others). Personally I'm very glad and grateful for the product of a tremendous amount of man hours I receive for free even if I too have ideas (that don't overlap perfectly with Go) about how a better language should be.
You misunderstand me, please re-read what I wrote. Nowhere did I call for the forking of Go. On the contrary, I specifically wrote that it was unrealistic to even envision it, since it would not work as long as Google sponsored more into its development that I could do with the fork, and as long as Google was perceived as a more stable future sponsor of said development.
What I wrote was that I, personally, would not use Go as long as its development was perceived to be controlled and paid by Google. Forking Go would not ameliorate this in any way, unless it was successful, which would be extremely unlikely.
Normally I don't reply in this kind of situations but I make an exception: 1 I think You misunderstand Go. 2 You complain about something you admit you don't know well. 3 You want the development to be more the way you want it without making any effort. My feeling is: Your internal feelings towards Go development are not relevant for many.
Your assertions are at odds with everything observable with regards to development of the Go language.
I freely admit that I do not know the intimate details of Go development, but my point is that they are almost irrelevant. Go is still perceived to be controlled by (and therefore is controlled by) Google. How much that is actually true is almost irrelevant until the perception changes. And with a name like “Go”, Google likely has no intention or wish for that perception to change.
I mean, can anyone claim that an internal developer at, say, Microsoft or Apple could develop programs in Go and have them become used for large parts of the internal company infrastructure without it becoming politically sensitive, just as if they had chosen, say, C? Until that happens, Go is not an obviously-neutral platform, and I therefore have no desire to use it.
> I freely admit that I do not know the intimate details of Go development ... How much that is actually true is almost irrelevant until the perception changes.
Since you freely admit your ignorance, can you please stop making uninformed statements that spread FUD about Go? Those of us in the Go community that invest our lives in this project don't appreciate your senseless negativity.
I don't think this is true of C. The original compilers were AT&T proprietary, surely? Hence the need for multiple competing implementations.
This is true of the very initial versions of C. My understanding is that it wasn’t really popular until the C compiler was freely given out to universities and later the world. Also, the book (The C Programming Language), effectively an easy-to-read language specification, contributed heavily to independent implementations, as the language filled a hitherto unfilled niche.
Have you actually observed the Go community and how Go has been developed over the last five years? What you are saying are fair concerns in the abstract, but I cannot reconcile them with the reality of how the Go project operates.
You also seem unusually hung up on its name. It's not like it's called Google Programming Language All Access.
I see from another comment that this release first happened, and then, afterward, the release was pushed to its “official” Git repository. This is not the way real open projects do releases, and instead indicates that the real development is done in-house and the code thrown over the wall.
A name is important, as it is a symbol. As long as the language is called “Go”, Google will always have power over it, no matter who actually does most of the work, and thus the fear will still be there. (Yes, “Go” is symbolically the same as “Google Programming Language All Access”. It’s as if Microsoft released something called “MSCode”.)
>This is not the way real open projects do releases, and instead indicates that the real development is done in-house and the code thrown over the wall
There are several projects that are free and open source, wildly available to every platform, hacked upon by hundreds of people, that still do cycle releases and development behind the back of most developers and only release full .tar.gz archives with sources after a milestone is reached.
I might be wrong but iirc bash is one of those projects (or at least was), I seem to recall people complaining about it during the shellshock issue. The GNU libc might be another but I'm not sure.
> I see from another comment that this release first happened, and then, afterward, the release was pushed to its “official” Git repository.
You misunderstand. The release was made from the official open source Mercurial repository, and later pushed to the official Git repository, because this release coincides with the project's migration from Mercurial to Git.
Every single change of this release was written in public, reviewed on public mailing lists, and committed to a public version control system. You are misinformed and spreading FUD. Please stop.
I see; thank you for clearing that up. I withdraw my comment about the release. I would edit it if I could.