How I explained REST to my wife | Hacker News Reader

How I explained REST to my wife

tomayko.com

199 pointsCodeChutney57 comments

I'm a woman, and I didn't find this patronizing or sexist, unlike some of the commenters on both the post comments and here.

I've gotten much more into web development in the past couple years, and this post reminds me of a handful of precious conversations I've had with incredibly smart people who took the time to bring together concretes and abstractions together in a way I could understand. I think this is a good post to share with other no-so-technical-people who you wish understood more about what you do.

Maybe I should write a post on "How I Explained Marketing to My Husband"

> I'm a woman, and I didn't find this patronizing or sexist, unlike some of the commenters on both the post comments and here.

I think his wife complained (in the comments) that she thought it made her sound stupid. Otherwise I agree, and I look forward to reading your post about marketing.

If I'm having a conversation with someone about a topic I know absolutely nothing about, I sound like an idiot. "What does that do?" "Who's that?"

Also, this 'conversation' was very 1-sided because the author was trying to prove a point. It really didn't matter what the wife said throughout but it let him explain further and use some real world examples for those who may not GET it.

Nice POST!

Great way to PUT it!

Feels like this guy has married ELIZA

That's his fault, not hers. You should see her version.

When I write conversations down it's hard to capture the spirit of the give and take. My straight-line write-down-the-meat approach always ends up devaluing the questions and ideas raised by others. My goal is to record what I learned from them, and I often come off sounding self-obsessed. I think it's an occupational hazard: all of us emphasize what we say in our own minds. Writing down conversations just uncomfortably highlights that fact.

My guess is that this conversation happened entirely in his head. His wife was in the room nodding and mumbling "uh huh" while he rattled off hacker-speak for way too long. He left off his wife's last line of: "Sorry, did you say something?"

Are you kidding? I'm sure his wife knows all about polymorphic verbs.

Please go on.

Really old, but a classic tale that needs to resurface every 6 months or so.

I am not aware of anyway to make an old submitted story resurface in HN. Probably some feature like this would be good to have?

that is actually a really good idea. if you could easily peruse previous "days" of hackernews. if an article is really worth reading once, it should be worth reading every few months.

"This day in Hacker News history"

I find that the people who are against REST, or don't understand why its neccessary, don't really understand the finer points of HTTP, either. This post really helps describe what it is about HTTP that makes REST obvious, and easy.

Rest is obvious and fairly easy. It solves a specific set of real-world problems.

On the other hand, that list of problems is a pretty small part of web application design. Solutions other than REST seem to work just as well.

When reading his intro to HTTP, I kept thinking of what Clay Shirky said,

"HTTP and HTML are the Whoopee Cushion and Joy Buzzer of Internet protocols, only comprehensible as elaborate practical jokes."

Read further if you're not familiar with this quote:

http://www.shirky.com/writings/evolve.html

As a noob to web coding, I appreciated the information although the format was a bit dragged out. Good share mrlebowski, I look forward to more of your info dude.

Thanks, I try to share good links that I come across learning stuff at my internship place :)

Remind me of this article: http://72.249.21.88/nonintersecting/2006/11/15/the-s-stands-...

This past summer I signed up for a Java Web Service course, which made my life a living hell trying to get SOAP, WSDL, and WS-* working. It also makes me appreciate REST a lot more.

So how does the machine, in these examples, request the machine understandable representation of the resource?

If the 'concept' is http://www.example.com/book/123, from this explanation, then there should be a way to get different representations of that via different verbs, right?

I guess this is where it always falls down for me, because there isn't a consistent way that all machine readable representations are flagged. Rather, we usually seem to define different URLs via extensions, for example, to define the different representations, which from this explanation seem like they should be reserved for different 'nouns' or 'concepts'.

Am I misunderstanding, or is that just the state of where we are at? And are there any plans to define those new 'verbs' at some point?

At the moment, generally you use GET and ask for a different Content-Type with an Accept header, or you say something like <link rel="alternate" type="application/rss+xml" href="foo.xml" />. In some cases you're expected to guess the alternate URL instead, maybe based on a document somewhere else, which is non-RESTian. However, it's pretty rare for someone to do what you suggest, which is to use new HTTP verbs in order to get different representations.

"are busy writing layers of complex specifications for doing this stuff in a different way that isn’t nearly as useful or eloquent"

What is that supposed to mean. HTTP is such a simple protocol that people need to invent technologies atop it because on its own it can't do much. lack of state, for example, so people came up with cookies and sessions and so on. These were necessary and an integral part of the web's evolution from the days back in the early 90's when people had 'home pages' of pictures of their cat.

I personally found the article badly written, overly long and full of passive aggressive attacks at what the web has become.

Cookies are part of HTTP.

He's referring to SOAP there I'm sure. And he's right. SOAP is awful.

That's just one protocol over HTTP, there are tons of others that are indeed useful and well written.

For me when I first encountered this particular post on REST a couple of years ago it was like an ah ha moment. SOAP is just one complex layer. Struts is another, actually a container in a container. Having messed around in J2ee land for while this article made me realize that most folks didn't know or didn't care what a stock J2ee container could do if one started thinking RESTfully.

Then we built a whole enterprise web application using just a stock J2ee container and RESTful style thinking. How much simpler! No extra frameworks to learn. The UI became completely decoupled from the services. The services are completely decoupled from the J2ee container, naturally. All the container does is resolve the nouns to a service name and give a call back to the appropriate verb method, doGet, doPut, doDelete.

In fairness we are not 100% RESTful there is some legacy stuff that relies on the session. However having done a few SOAP/WSDL services and tons of Struts actions or Spring "injections" I must say that REST is in my opinion a much cleaner way to code for web apps. Whenever I hear about some implementation needing a SOAP/WSDL service I run the other way as fast as possible. I won't even debate folks about it anymore, I just refuse to work on all that complexity.

As an aside I have seen here and there some frameworks the purport to use REST. I don't really get this. Why would somebody need and additional framework to make something RESTful? I have done it with J2ee and just plain Apache. I believe the web itself to be the framework, I could be way off in my thinking though, it wouldn't be the first time. Anybody have any clues about what a RESTful framework is for?

Mostly a "RESTful framework" just allows you to easily map verb/noun combinations to functions. You create a noun (an object) with the standard verbs as methods, and it does the other stuff. Pretty basic, but it makes life easier.

I am a designer so I never really got REST, but now I do. Definitely a must read.

As a designer you might enjoy the way I explain REST to people:

Think of the URL as part of the User Interface. REST offers conventions that make it easy to guess the URL of what you are looking for, just as you would guess where to click/navigate to get it.

I realize this only covers one aspect of REST, but I like it because I think the idea of an address as a resource which has a certain appearance (its URL) is quite powerful.

edit: I realize that rest could just as easily use meaningless identifiers, but it is often the case that meaningful identifiers are used such that the user can GET desired items by guessing the URL.

NO. NO.

REST is absolutely not about being able to guess URLs -- it's about never having to guess a URL. Repeat after me: Hypertext. As. The. Engine. Of. Application. State.

Clean, meaningful, organized / structured URLs are unrelated to REST, and their presence tends to lead to the predominant anti-pattern of 'API Specifications'. Leave that bullshit in SOAP's failure pile.

Uh ... I didn't say guessing state, I said guessing what is behind the URL. as in /users/greg vs /users/davis

rather than /u.exe?23q4234q34q34234234asdfa&query=true&xml=false&screen=home

The point I was trying to make is that with REST, URLs have meaning in and of themselves, not dependent on the presence/absence of a variety of other aspects of state embodied in the URL... The protocol (and REST verbs) is how you add that additional meaning... and a different URL thus means something different. Hence, for some universe of items, if one has an obvious meaning, the rest should.

If you are using the protocol in the right way then your resource will not change depending on which verb you are using.

It sounds like you're commenting on what REST is without ever having read Roy's dissertation. You can fix that, you know.

What was inaccurate about my statement?

I think the point that the previous posters were making is that pure REST should not rely on out-of-band information. The client should not have to guess or create URLs from knowledge they may have to look up in some other place. A pure REST API should be discoverable and subsequently navigable without having to rely on any other information not contained inside the request.

In practice this can increase the amount of requests required to get at the information you want, and also the size of each request, as it must then include information about the actions that can be taken on the resource. It could be argued that this constraint is more for the benefit of systems or programs that need to interact with the API without knowing anything but the initial URL, and not necessarily a programmer who can work these things out. owever, if a REST API uses this constraint, and application developers follow it, deep changes can be made to the URL structure of the API without breaking the applications that use it.

thanks for clarifying

It's not so much that it was inaccurate (although it was) as that it was missing the point.

fair enough... I guess the reason I have though of the idea is b/c of the resource focus of REST... where a URL means something... I realize that isn't the point of REST, but it's nice.

You said: "REST offers conventions that make it easy to guess the URL of what you are looking for, just as you would guess where to click/navigate to get it."

In your earlier comment it does seem as though you are talking about guessing the URL.

What the other comment was alluding to with "Hypertext. As. The. Engine. Of. Application. State." (or HATEOAS, one of the main principles of REST) is that your application should never have to guess a URL. The possible URLs your application can next GET/POST/PUT/DELETE to should be given in the response to the last action it took.

There is nothing in the original description of REST that says anything about URLs being human readable. That's a pleasant side-effect of a sensible design but isn't the main aim.

As the original post describes, the importance of REST is to make simpler for machines (not humans) to manipulate services on the web.

You should probably read a bit more about the principles of REST before confusing people with your own interpretation of what you think it is.

edit: mistakenly thought HN understood markdown. edit: mistakenly thought HN understood HTML.

ok thanks for clarifying

I still don't get it. There have been several things I never got or rather they were so obvious / intuitive to me there was nothing for me to "get". Still I spent years thinking I missed something cause so many others had such problems with them. Pointers & recursion come to mind.

There just doesn't seem that much to REST (that's not a bad thing). Lot's of hype, fud, misunderstanding built up around REST. Arguments over the proper way to apply REST to HTTP/Web. But, REST itself seems sort of duh, how else would you do it?

It's not magical rocket science, right?

Many (most?) novel innovations seem obvious and self-evident after the fact. Even rocket science (of course make a pointed tube, fins for stability, and _just_ the right fuel- how else would you do it?)

Have you read Roy's dissertation?

I read it and was surprised at how readable I found it (never mind being a dissertation).

http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm

What is also really good is Richardson and Ruby's "RESTful Web Services" book. They coined the term "Resource Oriented Architecture" to describe a specific REST design implementation and to side-step a religious war on what is and is not REST.

Here's how I would explain REST to a wife:

It's a made-up pile of bollox. It stands for "and now I'm going to refer to meaningless pseudo-science in an attempt to confuse you and bill more money to the project."

On second though, that probably explains why I don't have a wife.

How this can be posted again?

http://searchyc.com/How+I+explained+REST+to+my+wife

It still bothers me that web browsers don't really "do" REST without Javascript.

The title should read "how I explained REST to my trophy wife". Also, all the structured data stuff doesn't really belong in a discussion of REST, imho. SemTech researchers have been studying this problem for a long time, came up with RDF etc. Tim Berners-Lee was also you know, involved in those HTTP efforts, you can watch him discuss this topic here http://blog.ted.com/2009/03/tim_berners_lee_web.php . Not impressed with this REST article.

Reply on news.ycombinator.com