Google Go captures developers' imaginations | Hacker News Reader

Google Go captures developers' imaginations

thestandard.com

30 pointsmarketer16 comments

I'm past the point where I get particularly excited only any given language only because most great features get adapted by other languages before too long. As it is I don't see anything in Go that's significant enough to make me switch.

But what I think Google does need, and what Go could do for them, is to give them a standardized language. One of the things Microsoft does very well with C# is to make sure it works across the board. If you are a C# developer you have access to everything Microsoft does. You can still choose another language like Java, VB.Net, Python, etc... but C# is like a promise that it will work everywhere Microsoft is.

Google doesn't have that right now and I think it inhibits their efforts to recruit developers to their various efforts.

I think that's an interesting point, but wouldn't computer language polyglots actually be the sort of engineers that Google is trying to snare?

Yes, that's what I think.

Google seems to emphasize diversity and thinking outside the box (well at least from what I can see through the PR window and personal stories).

I think having one language do everything is not feasible. Some languages are good for system programming and are good for performance computing (I use C), some are good for what is not performance critical but are more concise and clear (I use Python for that). Go wants to do both, but from what I see in their benchmarks it is the worst of both worlds. I want it to be as fast as C, but even then I'll still fall back on Python probably for clarity and concision. So ideally I would want Python to be as fast as C -- not going to happen anytime soon.

What a company should promote is a common protocol for encoding and transmitting data i.e. protobuf -- that is a much better thing to have than looking for one perfect language that does it all.

Google has 3 common languages: C++, Java, and Python. I'm not really sure how they could get it down further. I guess they could go all-Java, but then we'd all be complaining about how Google does everything in Java (and a good many of their engineers would quit). Or they could do everything in Python + C++, but then all the Java engineers would quit.

If you just want to interact with Google's APIs from the outside, you can't really go wrong with Python. Almost all of them are RESTful anyway, any language with an HTTP client and XML or JSON parser should be able to work.

Syntax is important to me. Even if I know a language has certain great features, if the syntax itself doesn't appeal to me, it's hard for me to take it up. Go, Erlang, Scala and Ruby all fall in this category for me. Individually I appreciate certain aspects about each of them. But I don't care for their syntax. They repel rather than attract me.

I want something like Lispythonlang. One of you guys, go make that for me. I'll be taking a nap over here in the corner. K thanks bye.

I am looking forwards to a new system programming language - hopefully it matures well in the next few years. I still love C, but would love to be able to work in a little bit higher language. I have python for high level work. Will have to give go a try for my next tool app.

check out D, by Walter Bright of Digital Mars

it is older than Go

Bright is an old skool C/C++ compiler writer from the 80's. Seems to know his stuff. Co-authored one of my favorite computer games of all time too, Empire.

I looked at D. But it is lacking good "batteries" -- standard libraries (Phobos? Tango?).

For a language to be useful it has to have a very good standard library. Preferable one good standard library. That is probably the main reason why Python and Java became popular.

I picked Python over Ruby because it had a better standard library. Using Java after years of using C++ with various standard and not-so-standard libraries, was a breath of fresh air.

I liked D as a language, but I didn't like its librar(y|ies) so I discarded it.

I tink they are planning to fix this in the upcoming D2 release.

For this next release a lot of effort has gone into improving the Phobos library and I think this next release will also let Phobos and Tango work together.

Yet Go still remains as syntax clunky/ugly as ever.

For syntax, ugly yet familiar is more successful than obscure yet beautiful nine times out of ten.

How can you possibly substantiate that statement?

java:lisp

perl:python

Hey this is fun! Now you do one again.

Please. Python is so wildly C-like that it uses zero/non-zero for conditions.

You're either being willfully ignorant or insincere now. Allowing non-boolean values in boolean clauses originated with lisp back in 1960. Furthermore, GvR is on the record as saying that his main influence for python was ABC.

Reply on news.ycombinator.com