Re:Why Python is Powerful Enough for Google
by examancer on Monday January 27, 2014 @08:39PM
(#7134746)
Not a lot of meat to this article. "Ra ra, python" with no real comparison to anything else. You could rewrite the article for Ruby or javascript and you'd really just be changing the names of the tools/libraries, the points could remain the same.
Generators? Hello Enumerators in Ruby and using simple callbacks in node.
Speed? Hello V8 for JS. Hello vast speed improvements in MRI, C extensions, or leaning on the speed of the JVM for Ruby/JRuby.
Broad? Javascript is probably the most widely deployed language on the planet, and by many measures the most used. Ruby trails JS, and maybe Python too, but its still plenty popular. Not to mention the depth of libraries for both is pretty stunning.
Javascript is not javascript, ruby is not ruby. There are a number of vastly different and wildly popular JS implementations, and embedding JS in other applications is plenty common. Ruby has MRI (CRuby), Rubinius, JRuby (which is much more active than Jython), mruby (embedded ruby), and opal for compiling ruby into JS, not to mention the heavy ruby influences on CoffeeScript and Sugar.js.
The articles points are all valid, but none of them really demonstrate that Python is the answer to the article's thesis question: "You are looking for a job, which language should you learn?"
That said, Python is great. I like Python. Just not a big fan of poorly argued fan boy articles.
Re:Why Python is Powerful Enough for Google
by pathikrit on Monday January 27, 2014 @10:25PM
(#7135123)
I work at Google. Most of production code is C++/Java. Python is used for scripts, testing frameworks/test harnesses, internal tools like code review and bugs etc.
Re:Why Python is Powerful Enough for Google
by skybrian on Monday January 27, 2014 @08:22PM
(#7134674)
The decision to make Python an officially supported language inside Google was made a long time ago. It's anyone's guess whether they'd make the same decision today. If Go were around, perhaps they'd have chosen that instead?
Re:Why Python is Powerful Enough for Google
by aioprisan on Monday January 27, 2014 @08:18PM
(#7134646)
Google uses dozens of different languages in the services they offer. Heck, they even built a few languages themselves, like Go and Dart. Go is definitely used in production as well http://golang.org/doc/faq#Is_Google_using_go_internally.
Is there any data to suggest that Google writes a non-negligible number of apps in Python vs the dozens or so of other languages? I'm curious.
Re:Why Python is Powerful Enough for Google
by e12e on Monday January 27, 2014 @08:59PM
(#7134825)
They did start out their compute cloud [edit: I mean appengine] with python, and I think they started out offering mercurial hosting first on code.google.com? And they famously employ Guido. I also seem to recall something about reimplementing git in python for code.google.com, but I could be mistaken.
Re:Why Python is Powerful Enough for Google
by Bhel on Monday January 27, 2014 @09:48PM
(#7134995)
I like how they use a huge picture of a snake for the article, even when Python isn't named after the animal, but Monty Python.
More on topic: Even when Python is one of my favorite languages, this article is plain horrible. Biased, and without any sort of point to compare it to.
Python is great for many tasks, but it's just retarded to pretend that it's the best tool to go for everything.
Re:Why Python is Powerful Enough for Google
by slowdown on Tuesday January 28, 2014 @03:38AM
(#7136053)
This is the author's prejudiced opinion spun into an article. I could write the exact same thing for Ruby, Javascript, etc. PLus, if you are starting a company, the last thing that should matter is your tech stack. The first thing that should matter is your sales strategy.
Re:Why Python is Powerful Enough for Google
by vezzy-fnord on Monday January 27, 2014 @08:21PM
(#7134667)
I'm not against Python by any means. I like it and I use it fairly often.
However this article is very unconvincing.
The arguments summed up in order are basically:
* Python has generators.
* Python interpreters are getting faster.
* Argument from popularity.
* Holy shit, guys! Did you know that CPython isn't the only Python implementation, and that there's a difference between language specification and implementation? Mind = blown, amirite?
* It's so easy... scientists and mathematicians can use it? (Not to say that Python isn't easy, but to make it out like it'd be so impressive for a hard scientist to pick up programming is silly.)
Generators? Hello Enumerators in Ruby and using simple callbacks in node.
Speed? Hello V8 for JS. Hello vast speed improvements in MRI, C extensions, or leaning on the speed of the JVM for Ruby/JRuby.
Broad? Javascript is probably the most widely deployed language on the planet, and by many measures the most used. Ruby trails JS, and maybe Python too, but its still plenty popular. Not to mention the depth of libraries for both is pretty stunning.
Javascript is not javascript, ruby is not ruby. There are a number of vastly different and wildly popular JS implementations, and embedding JS in other applications is plenty common. Ruby has MRI (CRuby), Rubinius, JRuby (which is much more active than Jython), mruby (embedded ruby), and opal for compiling ruby into JS, not to mention the heavy ruby influences on CoffeeScript and Sugar.js.
The articles points are all valid, but none of them really demonstrate that Python is the answer to the article's thesis question: "You are looking for a job, which language should you learn?"
That said, Python is great. I like Python. Just not a big fan of poorly argued fan boy articles.
[edit2: According to http://google-styleguide.googlecode.com/svn/trunk/pyguide.ht... "Python is the main scripting language used at Google."]
"Employed", it appears: http://en.wikipedia.org/wiki/Guido_van_Rossum#Dropbox
I only just discovered this recently.
More on topic: Even when Python is one of my favorite languages, this article is plain horrible. Biased, and without any sort of point to compare it to.
Python is great for many tasks, but it's just retarded to pretend that it's the best tool to go for everything.
Yes, today it's all about me :-)
However this article is very unconvincing.
The arguments summed up in order are basically:
* Python has generators.
* Python interpreters are getting faster.
* Argument from popularity.
* Holy shit, guys! Did you know that CPython isn't the only Python implementation, and that there's a difference between language specification and implementation? Mind = blown, amirite?
* It's so easy... scientists and mathematicians can use it? (Not to say that Python isn't easy, but to make it out like it'd be so impressive for a hard scientist to pick up programming is silly.)