News for hackers. Stuff that matters.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by nahname on Wednesday November 20, 2013 @11:06AM (#6768806)
>TDD is huge in Java and Ruby ...40-50% ...JavaScript’s percentage of projects using a testing framework is considerably lower, coming in at 25%.

I doubt it is even that high. Sadly, many of the JS projects I look at have a tests folder and one test in it that usually amounts to:

equal("tests written", false)

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by litmus on Wednesday November 20, 2013 @11:11AM (#6768848)
I think the conclusion for Java is a stretch as well. Some sort of unit testing is probably standard in BigCo, TDD is not (I'm excluding the 'hot' trend-setting BigCos obviously).
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by byroot on Wednesday November 20, 2013 @10:18AM (#6768458)
Ruby ones are a bit pointless, rake, jquery-rails, sqlite3, uglifier, sass-rails and coffee-rails are all rails defaults.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by ddoolin on Wednesday November 20, 2013 @10:24AM (#6768496)
It would be more pointless if they were all exactly the same, but they're not. Clearly a good chunk of developers are interested in decoupling Rails from the defaults.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by weavie on Wednesday November 20, 2013 @10:58AM (#6768748)
It seems odd that there are more projects using Rails than Rake. What kind of Rails project would not use Rake?
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by ajasmin on Wednesday November 20, 2013 @10:13AM (#6768421)
What about the web and client-side JS?

From what I understand the list of JavaScript libraries is for node.js as it doesn't even include jQuery.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Wednesday November 20, 2013 @10:27AM (#6768514)
That is odd.

Grunt on Github: 6852 stars, 740 forks

jQuery on Github: 26,238 stars, 5,563 forks

And yet, "leaning towards those that have been favorited the most by developers". I'm confused on the results of what I think is their selection criteria.

Another thing, Javascript first appeared in early versions in 1995. Ruby on Rails appeared in 2004. But Javascript is still in the early days of it being a language? I wonder if that's in reference to recent growth in popularity and/or usage?

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by quarterto on Wednesday November 20, 2013 @10:38AM (#6768607)
Repeat after me: Ruby on Rails is not a language. Ruby on Rails is not a language. Ruby on Rails is not a language.

Ruby appeared in 1995.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Wednesday November 20, 2013 @03:27PM (#6770915)
Although true, does not alter my point.

Also, I don't see that I claimed Rails was a language, but I can see how it can seem I implied that.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by quarterto on Thursday November 21, 2013 @04:32AM (#6773820)
"Another thing, Ruby first appeared in early versions in 1995. Node.js appeared in 2009. But Ruby is still in the early days of it being a language? I wonder if that's in reference to recent growth in popularity and/or usage?"
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Thursday November 21, 2013 @10:17AM (#6775033)
I'm failing to understand your point.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by rohitv on Wednesday November 20, 2013 @10:55AM (#6768726)
I was confused too but, I think they analyzed 10,000 projects and the components they use. So, in conclusion a lot of JS projects use grunt.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Wednesday November 20, 2013 @03:29PM (#6770926)
A lot of JS projects require jQuery as well, or include it in some way.

I'm thinking my confusion is because, to me anyway, the description seems to suggest one thing while their list suggests something different.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by amouat on Wednesday November 20, 2013 @10:47AM (#6768674)
jquery is at number 19 on the full list with 197 of 10,000 projects using it.

https://docs.google.com/spreadsheet/ccc?key=0Alceaf46X4GPdHh...

However, I have no idea how they chose the projects or arrived at this number. Perhaps they missed all the projects that include jquery via CDN.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Wednesday November 20, 2013 @03:31PM (#6770943)
You're right, there it is. Although I see it at number 26.

I would be interested in more details how they came to that number as I'm somewhat confused by the list I see.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by warfangle on Wednesday November 20, 2013 @11:10AM (#6768843)
> As a result we see 50% more frameworks used in JavaScript than in Ruby and Java in the top 100, echoing that fact it’s still early days for the language.

I think this shows more how flexible Javascript is than that it's still early days (the language is 16 years old, only two years younger than Java). Many different frameworks have very different opinions; some are opinionated and some aren't; many of the libraries available to JS through, e.g., npm and bower are tiny little tools. That's the huge difference between a typical Javascript library and a typical Java library: many of the non-framework libraries included by Javascript projects are extremely small, modular, single-concern interfaces. So of course you'll see what seems like fragmenting.

But it's not, really.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by nivstein on Wednesday November 20, 2013 @11:23AM (#6768923)
While JS is old, node.js, which spans many projects on GitHub, is a relatively new architecture, which has only gained traction in the last couple of years. This stands in comparison to Java, which has a much higher degree of consolidation due to its maturity as a server-side language (this is also true to an extend with Ruby).
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by voidlogic on Wednesday November 20, 2013 @11:39AM (#6769042)
I'm not being pedantic for its own sake, but I have noticed many people using terms that have definitions for other things that already have terms defining them. Esp. the term architecture.

node.js is not an architecture, is a "software platform" for running a "language" (javascript) server-side.

An examples of a computer architectures are x86, amd64, MIPS, ARM, SPARC, POWER.

"For example, at a high level, computer architecture may be concerned with how the central processing unit (CPU) acts and how it uses computer memory." [http://en.wikipedia.org/wiki/Computer_architecture]

Now there also is a software architecture which is the abstract design of a system: "The word software architecture intuitively denotes the high level structures of a software system. It can be defined as the set of structures needed to reason about the software system, which comprise the software elements, the relations between them, and the properties of both elements and relations" [http://en.wikipedia.org/wiki/Software_architecture]

But node.js is neither of these.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by warfangle on Wednesday November 20, 2013 @01:15PM (#6769801)
Indeed. However, the typical accepted JS style, especially with NPM modules, is lots of tiny dependencies instead of a few large God dependencies.

Which will lead to what looks like fragmentation, but really isn't.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by jimktrains2 on Wednesday November 20, 2013 @10:16AM (#6768446)
I thought rails used rake, so how is the rake # < rails #?
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by ddoolin on Wednesday November 20, 2013 @10:23AM (#6768486)
It can be decoupled.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by tlunter on Wednesday November 20, 2013 @10:43AM (#6768636)
That's true, it can be decoupled. But not in the direction leaving Rake with less percentage than Rails. The authors could've deduped the results where Rake is being used by Rails, but I'm not sure they did.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by thekaleb on Wednesday November 20, 2013 @09:59AM (#6768331)
> Here are The Top 100 Libraries in Java, JS, & Ruby

Probably should've used the serial comma to make it more clear. At first I thought the top project language was Java and the second were projects that used both JS and Ruby together.

Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by tensafefrogs on Wednesday November 20, 2013 @10:24AM (#6768491)
This seems terribly biased towards more recent projects. Older projects might be used WAY more than the ones on Github, but because they started their lives on some other code hosting service, they will be left out of this evaluation.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by rubygnome on Wednesday November 20, 2013 @10:35AM (#6768571)
The results are biased towards favorited projects because it's a good indication of their perceived importance by the GitHub community.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by orenbarzilai on Wednesday November 20, 2013 @09:55AM (#6768314)
What about Python? pypi is a mess...
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by kmfrk on Wednesday November 20, 2013 @10:15AM (#6768437)
Ask the https://crate.io guys/gals to look into it. :)
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by orenbarzilai on Wednesday November 20, 2013 @10:17AM (#6768451)
Thanks for that, seems cool. Never heard of this before.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by kmfrk on Wednesday November 20, 2013 @10:23AM (#6768485)
Their auth system appears to still be hosed, though, so I don't know how much attention the project is getting.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by codegeek on Wednesday November 20, 2013 @10:07AM (#6768392)
Looks good. no love for python though ?
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by talmand on Wednesday November 20, 2013 @10:17AM (#6768454)
Funny, what is described as potential problems for Javascript I see as benefits.
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by jsvaughan on Wednesday November 20, 2013 @09:57AM (#6768327)
mockingito?
Re:We Analyzed 30K GitHub Projects: Top Libraries in Java, J... by ajanuary on Wednesday November 20, 2013 @10:03AM (#6768358)
It's a typo. The complete list linked to at the end has it as "mockito-all"