Elixir and OTP are really nice, but I'm frustrated by the type system, even with typespecs. I finally gave in and started learning Haskell. I would still choose Elixir/Phoenix for some web apps though.
Expect a long journey if you’re trying to get to the state where you can make a web app. I am making a conjecture here but from my limited but nontrivial experience it seems libraries in the applications ecosystem of haskell apply the most advanced features they know of to solve their problems. That would normally be fine, but they do not abstract away this complexity; due to the type system’s rigor, these choices are propagated to you as the consumer of the library because you have to consume the library in a way that satisfies whatever advanced type features they have. So the end result is that to get anything done like make a basic Yesod web app you have to be at an advanced level. I feel i benefitted from learning the structural and type theoretic aspects (monads etc) of Haskell but I also am glad i ditched it after some time because the investment just did not seem worth it anymore after a certain point. Anyway just my perspective, hope its helpful if you find similar difficulties. Right now I’m checking out Scala as an alternative, but i’ve just started so I can’t offer any comparison other than I know it is less rigorous.
> I am making a conjecture here but from my limited but nontrivial experience it seems libraries in the applications ecosystem of haskell apply the most advanced features they know of to solve their problems.
Apologies to others for the “me too” reply, but I wanted to let you know you aren’t alone, I also feel this way. It’s intellectual masturbation (without any reward/benefit), IMHO.
Haskell is a funny beast because I get a lot of joy from “conquering” the next level up (like a game) but in terms of productivity I’m not getting a lot done. OTOH in more pedestrian languages I’m using that brain BHP on getting stuff done. I really got into the Haskell thing a while back, going to meetups etc. but getting a job using Haskell is hard unless you’ve got experience already and want to take a pay cut.
Agreed - it certainly is rewarding in itself to get to the next level, but that's precisely it: the means become ends in themselves. I stopped using stuff like Arch Linux a while back for similar reasons.
Yep. Tools like Docker and Kubernetes on the other hand, I have found they give you a nice amount of leverage for much easier to understand concepts. In a similar "Workhorse" category I would put Git and Typescript. All of these require some effort to get to know, but they pay off nicely in terms of productivity.
Haskell is a pretty decent general purpose programming language, and if used well you can produce nice programs, but the problem is that the library you need will use some clever type system stuff and suddenly you are spending hours trying to understand how it all works.
For example I'd be happy doing a lot of stuff in the IO monad with some pure bits to the side where needed. Avoid the free monads, monad transformers, type classes and all that jazz. So use it like an imperative language for the most part.
The problem I have with Haskell for getting anything actually done is that nothing I try to do actually works. I mean in the basic "try out the tutorial to learn the thing and be able to get to the end without an error I don't have enough experience to debug" sense. Stack should help with this, but it doesn't when the versions you get (and function signatures of those versions) have changed since the docs were written. Cloud Haskell is an example of this: the tutorial docs are incomplete (the very first command they have you write is wrong, among other things), and they don't specify a stack version so the Hello World example doesn't build with the current snapshot and there's a really odd change to a public API I just can't get my head round.
I've had the same experience with Yesod and Scotty too, although not recently - every time I decide it's been long enough for the ecosystem to mature a bit more, I bounce off issues like this which I just don't get elsewhere, and decide to give it another couple of years.
I'm not familiar with Cloud Haskell, but I can attest that documentation and information availability in general have become a huge priority for me recently after getting burned by lack of it in recent years. This is actually very tricky to nail a sweet spot in, because it almost requires that a technology be currently popular, yet stable. There are long-lived Haskell libraries that have poor documentation and Haskell is rare enough that there aren't often answers to common pitfalls on StackOverflow etc. Ruby on Rails still has ongoing releases, but its popularity has declined and apparently so has its sources of community information beyond the official docs, which only cover so much. And then you have the immature, fast-moving JS ecosystem. Ironically for these reasons I could easily see myself choosing Java for projects in the near future; I don't like the language, but I've learned that the problems I have with Java are not my biggest problems as a professional engineer anymore.
Every time I see how many variants (and flags) the GHC has and I am like "feck no, I'll check it out again next year". But I think it's a culture thing in that community and it won't ever change.
That's what one of the things that's kind of making me give up on OCaml as well although there things aren't that bad.
Thank you for the input. Maybe I will never get to the point where I can comfortably make a web app, but learning the language has been a fun experience so far. I've heard good things about Scala and Akka.
I had to build a couple of Samsung TV apps, and that indeed looks like their IDE. It's Javascript (you can use HTML or Flash). Not much of a threat to Apple though, it is the worst platform I have ever developed for.
browserify-http can be used by OneJS as well but I don't see any point of emulating http module on client-side. I prefer using libraries supporting both platforms, such as boxcars; http://github.com/azer/boxcars
If you use node, I'd stick with browserify. RequireJS is nice if you want to load scripts asynchronously, but in reality your code is usually packed to one or two files so there is no need for the asynchronous part. Plus I dislike the RequireJS syntax, but that's personal taste.
Yes but as I said I don't need async loading of scripts, since all my code is packed in usually a single file. Therefore I don't need the RequireJS syntax, and I'd rather stick with a simple require syntax.
Help me, I'm still fumbling around in the web world und miss a lot.
Why would you even look at require.js if you don't have modular code/code of your own spread over a couple of files? This is for me the one single reason to look into it in the first place. To organize my own code better.
It's great thing that you use RequireJS to organize your code, and you should keep doing it. We use RequireJS at work and it works pretty well. However my personal preference is browserify, but use whatever works for you.
It took me more than four months after college to find my first job as an intern. Six months later, I got offered a better job so I quit. However during the transition I was offered some well paid freelance gigs so I didn't end up taking the second job. After 4 years of freelancing and moving out of a big city, I got tired of working alone so I joined a small team in a non-tech company. Six month later, I was getting bored so I looked for work in a major city. I found one startup that did something that sounded interesting, so I sent them an email even though they weren't advertising opened positions. An hour later I got a response, and two weeks later I was working for them. The past 3 months, I got offered 2 jobs and Facebook contacted me for the second time.
This isn't even the Silicon Valley -- I live in Europe. Sometimes I wonder if I'm smart or not. I know I probably wouldn't be able to answer all of these trick questions I hear about. I think I'm much better at abstracting concepts and seeing the whole picture. I love programming, but so far I haven't found a place where I can have as much fun as when I am creating software on my own, free of constraints and in control of everything. When I work on my own, every now and then I stop coding, sometimes for days, thinking, sketching, until I know exactly what I'm going to do and can explain why in plain English.
Anyways, I think younger programmers might enjoy taking time answering these kinds of question, but I'm the kind of person that wishes he didn't have to sleep so he could spend more time working on his own projects, so I don't bother. If you want me to be interested, you'd have to find a way to convince me that I would be working in an environment that feels like I'm working on my own project. I want to work with motivated people that care more about their product than getting funded. I don't want to work in an environment where the number of closed tickets is a metric for productivity.
One of these tricks is that you should fix 'the root of the problem' while a lot of the other tricks are hacks to ship the game on time. I'm a bit confused.
His patches comment was basically early on we realized the collision detection code was horribly broken so we just started patching every edge case we could find. That's practically an endless treadmill. On the other hand when there is a vary specific problem really late in the production cycle that has little do do with the rest of the game then you can just patch that specific problem and ship the code. When an audio driver you have no control over corrupts a single bit in your EXE that is the root problem not the symptom.
I'm glad that I see this now because I was going to launch a similar service in a month or two here in France. The wireframes are really, really similar.
I use redis a lot to store non critical data. For instance I store signup confirmation tokens in redis. The web app sends a message to RabbitMQ when a user signs up, then a background worker catches that message, creates an activation token, stores it in redis, and sends an email to the user. You can set an expiration time on keys too. It's convenient shared memory.
