Rest – A Haskell REST framework

http://engineering.silk.co/post/90354057868/announcing-rest-a-haskell-rest-framework

by bergmark • 12 years ago

127 47 12 years ago

engineering.silk.co

gberger• 12 years ago

Kind of a colliding name, huh?

dons• 12 years ago

This would become "haskell-rest" in most package systems.

lukasm• 12 years ago

Googlabilty of the name is really bad. Rename it to something else e.g. HRest

PMan74• 12 years ago

Or Resth, (unless somebody in the lisp community bagged that one already)

lynndylanhurley• 12 years ago

Hahahaha

btown• 12 years ago

Makes you realize that any possible name for a REST framework for Go is inherently an oxymoron...

anaphor• 12 years ago

Nobody thought of Hest? I admit it does sound kind of silly though, but it's got the 'H' at least, and it doesn't sound like Rest.

sfvisser4• 12 years ago

Really doesn't matter, the target audience will either use Hackage (Haskell's package database) or will prefix Haskell to their search term.

jamesbritt• 12 years ago

Or Rhest. (But even then it makes it hard to actually talk about in person.)

What motivates people to name projects after simple, common words?

Strikes me as a conceptual land-grab.

stdbrouw• 12 years ago

I'm more often annoyed by projects named after mythological creatures, vegetables or puns. They're fun for a while, but when I import something into a project, it helps if it's at least a little bit self-documenting – `import Rest` rather than `import Bazooka`, say.

tel2• 12 years ago

There's a little bit of that in Haskell right now. It's kind of OK with, say, `profunctors` which probably has everything anyone could ever imagine about profunctors in Haskell. Furthermore, as a mathematical concept there is not a lot of possible variability in that library.

Similarly, `lens` right now has pretty much everything under the sun about lenses, though some prefer the simpler lens packages like `lens-family`.

I kind of feel less confident about the choice around `rest` since there are many interpretations and implementation choices which may lead to a desire for competition here. But I'll reserve actual judgement until I use it. I wouldn't at all mind a comprehensive, provably correct webmachine-like implementation of REST called `rest`.

danidiaz• 12 years ago

I once wrote a Haskell library for Java Swing application automation. I thought of calling it HSwing, but to my surprise (and, eventually, horror) the domain was already taken.

codygman• 12 years ago

You could have named it Shwing, though that name is a bit exclusionary.

lclarkmichalek• 12 years ago

Interesting that they are using hs-src-exts, can the job not be done by Template Haskell?

sfvisser4• 12 years ago

It could, but would be extremely messy. We think hs-src-exts is really the right tool for the job here.

peaton• 12 years ago

I know this was just released, but are there any open source projects that currently use this?

bergmark3• 12 years ago

I haven't seen any projects except for rest-example https://github.com/silkapp/rest/tree/master/rest-example

But on the bright side we have been using this in production for a long time :-)

MaxGabriel• 12 years ago

Do you guys integrate this with Happstack/Wai/Snap at work, or do you use pure rest or something?

bergmark3• 12 years ago

We use both the rest-happstack and rest-snap backends, rest-wai is brand new so we haven't gotten around to using that one. You need to pick one of the backends to run an api, rest doesn't contain a web server.

thathonkey3• 12 years ago

"We've decided to call our REST framework - wait for it - Rest."

lelf• 12 years ago

This thread's queue for useless bikeshedders is full, sorry.

thathonkey3• 12 years ago

I disagree that the name for something like this is a trivial decision. Especially when considering the name for a framework which will need to be discoverable on Google to gain popularity. And even more especially when you choose to name the framework for working with a technological concept in one specific programming language the same thing (albeit different casing) as that technological concept itself.

How is that not a really important decision that should be considered practically? If I want to query Google for REST vs. Rest, or even converse with people about it, you don't think that will quickly become (needlessly) confusing assuming the framework becomes popular? And if you aren't expecting the framework to become popular, does that make it OK to choose such a lazy name?

But, hey, at least you got to namedrop the term "bikeshedding" along with your highly specious argument.

pstop• 12 years ago

Considering the key for any new user, when encountering problems when unfamiliar with the material, is to use a unique term to find answers on the web. Calling your framework after the technology is arrogant, foolish, and wholly unimaginitive.

dpeck• 12 years ago

Isn't writing out to the network a side effect?

Edit: Haskell folks are a bit touchy :) Just a joke, I use Xmonad, I'm cool with y'all.

arbsn• 12 years ago

Yes. Haskell provides excellent facilities for cleanly managing side-effectful code, like the IO monad.

dave809• 12 years ago

Managing side effects is one of Haskell's biggest selling points

dllthomas2• 12 years ago

Generally, it is the desired effect. In Haskell, it will be reified in an IO action, will appear in type signatures, and - as a value - can be manipulated and combined with similar values. It will only be a side effect if called with unsafePerformIO or similar, and the Haskell community will rightly condemn you for doing so (outside debugging) without a proof for why it's both okay and necessary in this circumstance - relying on side effects means you lose all kinds of guarantees that the type system usually affords you.

pestaa• 12 years ago

It may have been an innocent question, but it implied too much and it implied myths.

niix• 12 years ago

Very cool. I'm just getting into Haskell. Coming from the JS world, it's nice to be able to start to learn similar things that I do as a JS developer. My mind only thinks in the way of the web currently, so this helps my transition. :)

gambogi• 12 years ago

I don't have the expertise to dissect this code, so I will choose instead to criticize the name of the project. My opinion is valuable, relevant, and provokes further discussion.

autodidakto• 12 years ago

Instead of addressing the linked article, your post is insightful self-parody, with much sarcasm -- a style more suited to reddit than hn. Please remove it at once, preferably falling on your sword afterwards.

collyw• 12 years ago

Where are the HN rules? I always see comments referencing the style of comments that are encouraged / discouraged, without any actual reference.

seanmcdirmid• 12 years ago

There is a trend of minimalism in software system naming. I personally like it, SEO is overrated.

shawkinaw• 12 years ago

I was just thinking they missed a great opportunity to call it 'raskell'.

ilyagr2• 12 years ago

I always thought that what makes Haskell particularly readable are the types: to understand what a function does, it is often enough, and always helpful, to understand what type it takes to what types.

So, when the first type I see in a Tutorial is

    Resource IO (ReaderT Title IO) Title () Void

I am confused. When, in addition, the functions are named

     mkResourceReader :: ??? -> Resource IO (ReaderT Title IO) Title () Void

I lose hope that I'll ever understand what's going on. Granted, it was only recently that I understood what 'ReaderT Title IO' alone means.

I wonder, is there a simple explanation for what's going on, or is the problem at hand really so complicated that this is the easiest way to explain it?

For example, maybe there should be a shorter and clearer name for the type 'Resource a (ReaderT b a) b () Void'? What does one do with resources -- I can only see them plugged into the 'route' function.

At this point I started cursing that I can't find the Haddock documentation anywhere, and I'm too lazy to search through github.

I'm sorry if this comes across as unfriendly to the author of the library, but I hope it's helpful to see what a person who tries to understand what it does might go through. The more I look at it, the more I think that everything could be fixed with a better tutorial.

EDIT It's late at night, and I'm rambling, but perhaps it would be enough to explain in the tutorial what the composition

    route . mkResourceReader

does, and what its type is? It seems like that might be tractable.

sfvisser4• 12 years ago

I think your comments are valid and I mostly agree very much with this sentiment. Thank you for sharing.

But let's look at it differently.

This framework has been developed and shaped over a few years in a corporate environment by a small team of relatively experienced Haskell programmers. The code has been shaped and tweaked to specifically tailor all the needs we had at Silk for a effortless development of a very consistent REST API for our core product. The code solves a lot of problems at once.

We never developed this framework as a minimal library for public use. But, because it solves and actual problem for us and is very general purpose we felt obliged to open source it. This might be very useful for other as well.

One of incentives for a company to open source an internally used framework is to get community feedback that might end up shaping the software to be simpler, more generally applicable and more battle tested. That's why your critique is useful and exactly one of the reasons to opens source.

To go more in the why the types for a Resource are what they are right now:

- In a real world example your API probably doesn't run in IO, but in some monadic context that allows acces to you database, configuration settings, log server, S3 buckets etc.

- Resources contain a schema type that statically enforce the invariant that no routes will ever overlap in the server.

- A single resource, or multiple resources, can be identified by multiple identifiers (e.g. a user by uuid, e-mail, or last name). We require a total mapping from identifier to resource action.

- While stepping deeper into an API (path segment by path segment) we can learn more about our context and pass information down to sub-resources. We do this by extending the context, in the simplest case with a reader.

- The `route` function builds up an existential. So it looks like the type of the Resource is irrelevant and could be one simple type synonym, but it isn't. It's different every time and carries a lot of static information about the resource used at runtime, and for code and documentation generation.

We'd love to hear how we can simplify this.

ilyagr2• 12 years ago

There's nothing wrong with what you are doing. There is no law that states that all open source libraries in the world should be perfectly documented, and have a set of tutorials that would gently teach anyone as much as he wants about how to use it. Writing documentation and/or tutorials is hard work, and the larger your target audience is, the harder and more time-consuming it is.

I don't think your library will be generally useful to a large audience until you or someone else writes better documentation for it. But you don't have to target a large audience. It's quite likely that when someone else has the same problem, it would be faster for them to reuse your library than to write their own. They might even document the process.

I don't think I'm qualified to give any specific advice, but here's some general suggestions which may or may not be to the point. If you are sufficiently motivated, you should decide who your target audience is, and write enough documentation for them. For example, for an advanced audience, it might be useful to say a few words the general design of your library, explain which problems your library does solve and (more importantly) what the limitations of the design are, perhaps outline anything in the design that is strange or unexpected, but useful from your experience. The most important thing for them to know quickly is whether your library can actually solve their problem (which is probably slightly different from the problem you solved), whether it is actually faster to use your library than to roll their own. If they really need to, they'll figure out the rest. The end result might be closer in style to a CS paper than a tutorial.

Oh, and one thing you really should fix: it is annoying that the haddock documentation is not available on Hackage at the moment, and probably not searchable on Hoogle as well.

Again, there's nothing wrong with your library, you are doing a good thing open sourcing it, and any time you put in describing it and explaining it to people is valuable. Just keep doing more of what you are already doing.

pepijndevos• 12 years ago

Nice! I worked with this project when it was not yet public, and it's good. A lot easier to use than what the monads might lead you to think. The generated API clients are great too. Of course the Haskell client uses the same types as the server, so you get type errors in the client if the API changed.

mark_l_watson2• 12 years ago

I have a Haskell project that exposes some NLP experiments in a Yesod web app. I would also like to provide a REST interface, so this looks good.

That said, I found nothing in the documentation for using this with Yesod and the developers seem to best support happstack.

I am a newbie at using Haskell for web apps, so I need to dig into the possibility of mixing happstack and Yesod in one app.

mark_l_watson2• 12 years ago

While this looks like a cool project, since I have already started using (and like) Yesod, I am probably better off using Yesod's REST support (http://www.yesodweb.com/book/restful-content)

Rest – A Haskell REST framework

Related Stories