Rosetta Code | Hacker News Reader

Rosetta Code

rosettacode.org

73 pointsbananacurve12 comments

I think this is a very good way to get a quick idea of how a new language works. No theory, no blabla, just the code in a language you know well side by side with the new one.

It actually works well with spoken languages: I've done that by buying the same book in my native language and a new one for which I know only the basics, with great success (and perseverance).

Your method of learning spoken languages is pretty much the same one that James Murray [1], the editor of the Oxford English Dictionary [2], used to learn new languages, which he did many times. He used translations of the bible.

[1] http://en.wikipedia.org/wiki/James_Murray_(lexicographer) [2] http://en.wikipedia.org/wiki/Oxford_English_Dictionary

I am a big proponent of this method for programming languages but this is also how French was basically taught to me in school and it just did not work. The difference? You can practice coding alone... A Hunan language is much more difficult to do that with.

Rosetta Code proved very useful in an experiment to see how typos affect code written in various programming languages http://www.spinellis.gr/blog/20121205/.

This is interesting and helpful (provided that the fuzzing does not have some language bias), I would like to see more languages in that list though - for example, Go (it should do very well).

This has been a cool resource, but its comparisons of tasks can be profoundly misleading to novices...not intentionally, just because novices don't know how to really judge things. Take, for example, its section on Web scraping:

http://rosettacode.org/wiki/Web_scraping

First of all, web scraping is a complex task. Not just at the low-level code, but conceptually...web scraping means different things to different people and this shows up quite frequently even in HN discussions.

In any case, the Rosetta example does kind of the right thing by limiting it to a very simple operation...but even the operation itself is misleading (parsing HTML with regexes?). And even in this idealized operation, you have very different opinions on what it entails. The Ruby one looks correct to me, and it looks very concise compared to other languages, but only because, unlike the CoffeeScript example, there is no error handling.

The real problem is that when you're a beginner, you have no concept of how programming is best done in small steps...Web scraping is simply a chaining of much smaller steps, such as string manipulation, parsing, options setting, etc. It's not a "feature" of a language, yet if you were to make a judgment on Rosetta (and who else but a beginner would really use Rosetta to compare languages by webscraping), you're going to have an unhelpful perception of programming languages and programming in general.

That said, it's a great resource when the task is low-level and well-defined. But what that actually means obviously differs for people and languages. Hence, opinionated frameworks and the like

As long as people go to that site with the intention to learn languages, and not learn the concepts involved, they'll quickly realize that like wikipedia, the quality of the submissions is sometimes lacking. In fact, a lot of pages have multiple versions of the same program, say, in ruby, with an imperative style and a functional style, or using different language features and degrees of idomatic...ness.

Greatly helped me understanding huffman.

The thing I find most interesting about rosetta code, is not the quantity of languages, but the actual algorithms which are used. There are many great algorithms out there who can be done in very short bits of code, which are great for educational purposes.

It would be better if the solutions were vetted to be like for like. For example in http://rosettacode.org/wiki/99_Bottles_of_Beer, some of the solutions add special logic for handling case 1 (singular) and 0 (go to the store), while other solutions do not (arguably correctly, since the spec didn't request it).

Having different solutions implement different algorithms doesn't really help somebody just trying to get a sense of the language.

I've been using this site for quite a while when I first made my jump from OO and imperative programming to functional. The side-by-side comparison was very helpful. They also have code for just a ton of algorithms in general.

There is also http://en.literateprograms.org

I contributed to the RREF implementation when I was in high school on here for PHP.

Reply on news.ycombinator.com