Hacker News Reader

Comment by teddyh | Hacker News Reader

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.

View on HN

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.

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.

Reply on news.ycombinator.com