awesome, didn't know about this. I watched a few of eric meijer's videos on youtube a bit ago but got distracted by his tiedie shirts.

Outside of a few people on HN saying they use it for some secret finance applications, this is pretty much what I've read everywhere about haskell. I'm mostly interested because I still don't know what a monad is after watching some douglas crockford talk. I just want to know what the cool kids are doing

Cool, bookmarked

Yeah, some hedge funds use it, but as far as I can tell, it's not so much they care specifiically for haskell or need haskell to do what they do but more that hedge funds like to hire ultra-smart, mathematical types and those types tend to gravitate towards haskell. Also for novel financial modeling and exploratory work, haskell's ecosystem weaknesses aren't as relevant.

Also, this is really good:

http://en.wikibooks.org/wiki/Write_Y...me_in_48_Hours

lol.

i dont know what the cool kids are doing, but they are certainly not sitting in front of a computer writing haskell

snorting blow off a hooker while writing haskell with one thumb on an iwatch?

You linked this one a long time ago. Been meaning to do it for the longest time.

***

Phone interview for Jr DBA tomorrow. I wish I could say I was super excited or super nervous, but I'm not anything.

I have no idea what to expect. I'm pretty confident but I know I can fall flat on my face pretty easily. I've been looking online for practice questions, what this job would entail, and brushing up on things I can never seem to recall off the top of my head.

The last place asked what I thought were fairly basic questions, but mostly focused on queries. I wasn't sure how well I presented myself though.

How did you know what Ron and Cally were doing this night.

What do you guys think of nodejs and why are you not learning it?
Its like coolest toy to play with right now for me.

gl daveT

Because I've been playing with R lately.

http://mazamascience.com/Classes/PWFSL_2014/

Longshot here, but does anyone have experience combining AngularJS with ASP.NET MVC? I'm also interested in general opinions on Angular from anyone who has worked on it.

Background on me: I worked as a programmer 2004-2006 or so primarily in ASP.NET webforms, then quit and played pro poker for years. I got a job again about 7 months ago and am still trying to get my head around the various innovations in the field since then. I've spent some time just recently learning AngularJS. I really like what I've seen of it, though I haven't used it in practice. The company I'm in is dedicated to ASP.NET MVC, so any use of Angular would be an extension to that. It seems like in some ways they might complement each other (Angular being client side and ASP.NET mostly server-side) but might also tread on each other's toes a bit (both have built-in client-side validation, ASP.NET model binding seems incongruous with Angular's JSON-based data structures, etc).

A good book by Brad Dayley, is Node.js, MongoDB and AngularJS Web Development

My take:

Angular solves a lot of problems. The experience of building with Angular can be one where tasks that were previously very difficult are suddenly trivial. Angular definitely fits into a .Net SPA architecture, but it complements Web API much moreso than MVC. But being "dedicated to ASP.NET MVC" is really going to hold a company back, imo.

The thing is, Facebook and Google are basically competing in this space, and Facebook is winning the battle of ideas hands-down with React. Angular solves problems, but creates a bunch more in the process, and continues down the road of messy mutable state that can be "easy" if done right, but never simple. With React and immutable-js, we're starting to see how it's possible to actually sleep at night when you've pushed state to the edges of your frontend systems.

Now, React has the right idea, but it's still stuck in javascript-land, which everyone agrees is not a happy place. Add ClojureScript to the mix, with interfaces to React like Om, and you're getting to the promised land. ClojureScript has built-in immutable data structures, so a Flux-like unidirectional data flow just falls into place without even trying. Functional UI programming, with a beautifully hand-crafted language like Clojure(Script) is the absolute nuts. Even the people at Facebook who invented React and build stuff with it daily have stated publicly that they'd switch to ClojureScript in a heartbeat if they could.

Meanwhile, the Angular project seems to be reeling. Angular 2.0 is coming, and they've supposedly been working to address some of the inherent flaws, but it's getting turbulent (link).

chris,

as someone who invested a good amount of time learning angular, my advice is: don't. i would never go back to it. it's a bloated, over-engineered framework.

if you are coming back to web development and looking to learn the best new js technologies, imo you are much better off with facebook's react.js (if you want a big project with a big community) or the even slimmer, but similarly architected, mithril.js (http://lhorie.github.io/mithril/), my personal favorite. you can learn it in a half a day to a day. the design philosophy is simple and beautiful.

Reeling isn't quite the word I'd use:

http://www.google.com/trends/explore...3%2023m&cmpt=q

http://en.wikipedia.org/wiki/Greenspun's_tenth_rule

http://www.joelonsoftware.com/articl...000000020.html

http://www.joelonsoftware.com/items/2006/12/09.html

This is pretty much our company tech stack for all the web stuff, though I'd say we really just use ASP.NET MVC to host a mostly static angular page and for minification/bundling and such, with all application logic in Web API and AngularJS. It's great, no complaints.

so.... you are saying that simplicity is not an important value in software design and any criticism of anything as bloatware is impractical idealism?

i've used all three of these frameworks, and angular is awful compared to the other two. ymmv, of course, but you can't have this argument on the level of general principles. to be clear: when i say bloated and lacking in simplicity, i am primarily referring to the learning curve and the degree to which you are tied the framework. angular is a framework-with-a-capital F, and when you use it you do things the angular way and you tie yourself to the framework. react and mithril are closer to libraries than frameworks.

if you've used all of these frameworks and disagree with opinion, we can have a conversation. but suggesting that i'm likely to be wrong because i'm using words like simplicity and bloated is silly.

EDIT: also, joel's article against simplicity is mostly an exercise in semantic gymnastics. and note he concedes: "If you're using the term "simplicity" to refer to a product in which the user model corresponds closely to the program model, so the product is easy to use, fine, more power to ya."

funny, the view I hear from the ember side is that angular doesn't provide enough structure.

my criticisms of angular probably apply doubly to ember. in fairness i have not used ember other than cursory investigation, but even more than angular it's a Big Brother Will Help You, just do everything how we tell you Framework.

EDIT: what they probably mean is that, similar to python, there is more clearly a "correct ember way" to accomplish any particular task, whereas with angular there are multiple ways to attack something. which is a reasonable point of view. if you are going to go the full on framework route, you may as well at least have an opinionated one that guides you with a firm hand.

All this angular talk is kind of moot. We know there are breaking changes coming in 2.0 and it sounds like it will mean essentially learning a new framework.

imo there is no sense adopting angular now since there are plenty of good options that will support what you build better and longer

Well, then, talk about something concrete.

Also, the primary virtue of simplicity is that it has better survival characteristics:

http://www.jwz.org/doc/worse-is-better.html

Given that AngularJS is not just winning the war, but completely obliterating the competition, I'm not sure why you're going on about simplicity.

From:

http://lhorie.github.io/mithril-blog...m-angular.html

"This means that instead of having to retrofit control flow into a declarative syntax that was never meant to support it, we simply take advantage of the standard DOM API that ships with all browsers to create and update the HTML from Javascript instead, and let Javascript do the job of the control flow, variable shadowing, recursiveness, etc. Use the right tool for the right job, etc."

Followed by:

https://github.com/insin/msx/blob/ma...sx/example.jsx

https://github.com/mozilla/sweet.js/wiki/Example-macros

Standard indeed.

How would you use Mithril with libraries that control DOM rendering? AngularJS's declarative syntax works seamlessly even within HTML rendered by third-party libraries.

Some of us have to deliver solutions today.

As for AngularJS 2.0,

http://www.reddit.com/r/angularjs/co...r_0_live_post/

"Please let us build 2.0 first. 'It's very hard to build a bridge to a city that doesn't exist.' - Brad Green"

"There will be a migration path, we just don't know what it will look like."

Honestly I don't know what the two links are intended to show so I'm not sure how to respond.

I have not encountered the problem of using Mithril together with libraries that control DOM rendering, so I don't know. There may be a way, there may not.

That, for me, is very far down on the priority list. At the top of the priority list is the ability to use POJOs to do all my domain modeling -- and to build those and test them as if the framework I am going to use does not exist. That's a requirement for me. Right after that is being able to accomplish easily the tasks the framework supposedly helps me with. That means spending very little with docs, framework-specific concepts and terminology (directives, transclusion, etc), and elaborate video tutorials or books to do something that should be simple and basic.

Your values may differ.

Right and if you're already using or know angular, then great.

But with angulars existing flaws, steep learning curve, and the prospect of another steep learning curve likely coming with v2 why would you recommend someone learn angular?

What else would you learn? What does AngularJS 2.0 have to do with anything? AngularJS 2.0 doesn't obsolete AngularJS 1.x any more than it obsoletes any other JS library that exists today. If it's successful, it will make every current framework obsolete - if it's not successful, then AngularJS 1.x is still more popular than every other JS MVC framework by a wide margin. Users of AngularJS 1.x are no more forced to use AngularJS 2.0 than users of Ember.js or Backbone or whatever. If AngularJS 2.0 is successful, it's almost certain that the migration path will be easier coming from AngularJS 1.x than from any other framework.

And no, concepts in AngularJS aren't arbitrary ideas that are useless elsewhere. If you fully internalize the design decisions they made, read through the source code, etc, it will make you a better programmer.

Those aren't my links, they are from the Mithril/AngularJS comparison put up by the Mithril dude, as examples of "benefits" of using Mithril.


Has anyone? I'm seeing zero evidence of Mithril being used for anything. There are 17 questions on this as opposed to 64K+ for AngularJS. If you aren't already mainstream, you better be much better than popular existing frameworks and better provide a great interop story. I can't find anything Mithril does better than AngularJS, other than being small and potentially faster, provided that you do more manual plumbing.


I learned AngularJS in 2012 before there were books or video tutorials (that I knew of) and before their documentation was even decent and it took me a few days to learn it well enough to write a fairly complex application. Complaints about AngularJS at this point sound like complaints about Java and .NET standard libraries. You don't have to use features you don't like, but they are there because somebody needed them. Pet little frameworks are unlikely to become popular until they add all these missing features, by which point much simplicity and elegance may be lost.

Also, it's baffling to me that you're in essence advocating learning esoteric frameworks that are easy to pick up and already fit your mental model. Esoteric frameworks are not going to be useful from any kind of career or business standpoint. Why bother, unless they force you to change your mental model and learn something new? AngularJS was worth learning for me even if it wasn't going to be popular precisely because it introduced interesting concepts. What's the point of learning Mithril, with its 17 stackoverflow questions, if you're not picking up something novel along the way? It's like somebody who already knows Java, C, C++, Fortran, Cobol, Basic and Pascal evaluating Algol, Prolog, Scheme and Haskell over the weekend and deciding to learn Algol because it doesn't introduce language-specific concepts and terminology (monads, applicative functors, continuations, hygienic macros, unification, etc) and doesn't make simple things hard.

I think monads would seem less mystifying if people didn't generally jump into their mathematical underpinnings and how to write them instead of just explaining how they're used and what they're for. e.g. I like Erik Meijer's definition of Monad: "The thing that allows do notation to work"