Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The New Haskell.org

haskell.org

Post reply on HN
Print view View original

61–70 of 110 posts

Re: The New Haskell.org

#61

Post by chrisdone »

scottjad wrote: post #35

This redesign appears to take the position: the first page on haskell.org is for people who are brand new to the language, not for people who are already learning the language or currently use the language. I don't like that choice. I think the page should be trying to be the most useful for users of the language with a link to an About or Getting Started page that is geared toward new users. To me, a user of the lan…

> This redesign appears to take the position: the first page on haskell.org is for people who are brand new to the language, not for people who are already learning the language or currently use the language.

Correct! See here for a complete report on the motivations behind each design decision: http://chrisdone.com/posts/haskell-lang#the-audience

> To me, a user of the language visits the site dozens or hundreds of times over the course of a career, while the average brand new user only visits it (I'm guessing here) 1.1 times.

In practice this isn't true—it's generally agreed in the Haskell community: nobody goes to the home page, they go to either Hackage, Hoogle or google something and reach a page on the Wiki.

Re: The New Haskell.org

#62

Post by chrisdone »

doctorstupid wrote: post #48

"Hi there, [name]! That's a pretty name. Honest. You're getting the hang of this!" I fear that this style of childish and somewhat condescending tutorial content is becoming fashionable. Was it the best fit for the audience?

The tutorial is 5 years old, so I guess that makes me a hipster tutorial writer because I was doing it back when it wasn't fashionable? =p

Re: The New Haskell.org

#63

Post by Ixiaus »

This is a step backwards imho. The fit and finish of this site betrays much about the language and its amazing community.

This feels amateurish, with enough people complaining about the interactive repl and lacking information, this should have gone through more iterations and feedback loops with people outside the Haskell community.

Re: The New Haskell.org

#64

Post by raiph »

chrisdotcode wrote: post #14

This new, awesome site is using the framework Yesod[0], which is kinda like the Rails of the Haskell world, only it can do rad stuff like prevent XSS and 404s at compile time . Yes, you read that right. [0] http://www.yesodweb.com/

It sounds like the feature you're highlighting here is static checking at compile-time of how routes are going to resolve at run-time. Is that right?

Afaict Perl 6 covers this case:

http://jnthn.net/papers/2015-fosdem-static-dynamic.pdf#page=...

Or am I missing something?

Re: The New Haskell.org

#65

Post by dmit »

grey-area wrote: post #33
rimantas wrote: post #24

Earlier quoted context omitted.

Tell me more about preventing 404 at compile time. I call http://www.yesodweb.com/bullshit

Presumably the idea is that you can prevent your site having broken internal links by constructing links with data types and checking the links when creating them at compile time. You could actually do something similar in most languages, even dynamic ones, so I don't see what is specific to haskell about it, e.g. rails has link_to and friends which require you to have a route set up in order to link to it (though no…

Let's be fair here, while you can detect invalid internal links in dynamic languages at runtime, the end result is basically the same for the user - a 404 or a 503 are equally bad in this case.

That's the advantage of compiled languages over dynamic ones in these kinds of scenarios: you sacrifice immediate feedback during development to gain a guarantee from the compiler that all invariants you embedded in the application uphold. And the more powerful the language/compiler combination, the more compile-time guarantees you can specify.

Re: The New Haskell.org

#66

Post by chadaustin »

bonn1 wrote: post #8

Out of curiosity: Who has written Haskell & deployed to production in the last 24h and if yes for what kind of app?

Not in the last 24h, but I pushed to production on Friday. We use Haskell at IMVU to power some of our web services and persistent stateful servers like the matchmaker.

http://engineering.imvu.com/2014/03/24/what-its-like-to-use-...

Re: The New Haskell.org

#67

Post by iopq »

λ do line > л

mueval-core: Imports.hs: removeLink: does not exist (No such file or directory) ExitFailure 1

What did I do wrong?

Re: The New Haskell.org

#68

Post by shangaslammi »

skrebbel wrote: post #16

I'm new to Haskell, and I like this page. So, wanting to get a feel for it, I tried typing the sieve example from the top of the page into the "Type Haskell expressions in here." box right beneath it. First, I got :1:8: parse error on input `=' Hm so apparently definitions aren't expressions? Or something? I decided to remove the `primes = ` part, which made my browser really slow and had no result. I'm assuming it w…

Yeah, the online REPL is definitely not very useful or intuitive yet.

As it says, you can evaluate a single line Haskell "expression" and top-level declarations (like the sieve example) are not expressions. In addition, the sieve example declares an infinite list so there's no sensible way to print it.

You can fix all the above by using a let-expression on single line and only evaluating a finite part of the list using the 'take' function.

    let { primes = sieve [2..]; sieve (p:xs) = p : sieve [x | x

Re: The New Haskell.org

#70

Post by nightski »

cbd1984 wrote: post #40
skrebbel wrote: post #16

I'm new to Haskell, and I like this page. So, wanting to get a feel for it, I tried typing the sieve example from the top of the page into the "Type Haskell expressions in here." box right beneath it. First, I got :1:8: parse error on input `=' Hm so apparently definitions aren't expressions? Or something? I decided to remove the `primes = ` part, which made my browser really slow and had no result. I'm assuming it w…

The biggest problem isn't what others have said. The biggest problem is that the try box doesn't let you define your own functions or variables. Therefore, even if you were an old hand at GHC and GHCi and so on, you'd still be unable to run the code it shows at the top of the page. This is probably bad website design. Especially since the website still lets randoms burn themselves by entering non-terminating code tha…

I agree with the problems but am curious how you would solve the non-terminating code problem?
Post reply on HN