Why Google's Go Might Be A No-Go

After news of Google's Go language surfaced, I went to my programmer friend for some additional perspective on Google's new experiment. He wasn't impressed -- and actually, neither was I. We had different reasons.

My friend's cynicism stemmed from his own experiences as a programmer, where he has seen languages come and seen them go (pun not intended), and been uniformly unimpressed with the parade that's clomped past his window. He had his own distaste for Clojure ("LISP in a VM", as he put it), F# (OCAML for .NET runtime, nothing more), and so on -- not because they're bad languages, but because the effort put into creating a whole new language doesn't seem to have been justified. He looked at Go and saw very little reason to be interested in it over C# or Scala -- or for that matter, Java or even Perl. Languages are not in short supply.

"I don't have anything against a programming language devised experimentally," was how he put it. "But what's the experiment?"

Is the experiment creating a new programming language? I didn't think so. I went a step further. What Google is trying to do, I thought, is use open source, and Google's existing programmer's fanbase, to create a community around the language so that it can be expanded for them without them having to do the lion's share of the work.

Part of the reason I believed this was because the language as it currently stands is such a stub that it's difficult to accomplish anything of scale with it. Perhaps they think in the long run something like Perl's CPAN can emerge, a library of work contributed by those who use the language.

My friend's cynicism is rubbing off on me, I think. That and I have grown suspicious of Google as a benevolent entity. Like any company that uses open source, they'll use it in a way that benefits them first, and there's little point in pretending such things never happens. (As my colleague Thomas Claburn noted, even the name of the language is turning into a subject of dispute.)

Whatever the benefits Go may have has a language, I suspect the real problems programmers face today aren't going to be solved my new programming languages or environments. The biggest issues I see are that many of the platforms -- especially the Web -- are lousy delivery targets. Writing applications for the Web that serve up anything more than static pages requires JavaScript, and JavaScript's problem is not that it's a bad language but that its behavior is completely at the mercy of whatever browser vendor your clients have decided to run with.

If Go has any real benefits other than being a Google-branded creation and being open source (under the BSD license, no less), it's far from obvious right now.

InformationWeek and Dr. Dobb's have published an in-depth report on how Web application development is moving to online platforms. Download the report here (registration required).

Twitter: Me | InformationWeek
Facebook: InformationWeek