Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

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.



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.


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.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: