I think your posts are all trying to make different points.

What I was saying is that every idiot on earth can program in Python these days. There is no badge of honor to say you program in Python. Every free and non-free resource teaches you Python to start with. Now you can take a 101 class and you "know" how to "program" in Python.

ad-hominem much? here's the background on Jack Dorsey: http://en.wikipedia.org/wiki/Circle-ellipse_problem

This is an absurd argument to make. The fact they are both on the JVM has absolutely nothing at all to do with the run-time of the languages. The fact they are on the same run-time, and on a run-time optimized for Java specifically, means that by definition, any language written on the JVM, whether it is JRuby, Scala, or Clojure, will be slower than Java in many cases and will never be able to out-perform native Java code. This is exactly like arguing that Python is slower than C even though it is ported on the same archetecture, and OMG, is compiles to byte-code, so why is Python slower than C?

You didn't read the follow-up to the conversation. Candybar didn't do the initial cut-down.

"Best Tool for the Job" usually obscures the fact that you are simply using the language / tool you are comfortable using. No one actually ever does the research to discover the "Best Tool" if such a thing even exists. Imagine the person who is normally used to using PHP or Python: someone who is very intelligent sees their project and says the "Best Tool" is Gambit Scheme. Do you honestly believe someone is going to learn Scheme and work on tweaking the pre-compiled C pointers?

Yes and No. I'm not sure what the initial motives are, but I suspect that many times it was to just create a language they enjoyed working with. Matz doesn't do Ruby full-time, even.

No, I simply admonished the poster for pounding the drum of religion and then backing up his religion with half-baked truths and outright lies.

I then pointed out that, despite his world-view, many people look down upon him as he looks down upon others. I was not saying I look down and honestly, I'm partly ashamed of using Lisp because the language is rarer and the community at large is irritating. Using Lisp is like wearing a badge that says "I'm really ****ing weird and proud of it."

When did Twitter switch from Scala to Java? Anyway, I don't think anyone can argue against getting away from RoR. I haven't see that whale in quite sometime.

Really? It still seems pretty popular to me.

Disagree. It's a badge of honor to use Python over some other languages for a given problem. It's the best tool for the job for many jobs and many definitions of best.

Interesting. I have a lot of experience doing this. LOL I need a raise. Multi-threaded apps might be a good thread (pardon the pun) topic. Personally speaking, architecture/design coupled with an iterative type development process are the essential aspects in being successful.

It's only hard to do if you pick crappy tools. Functional/non mutable state languages solve about 50%+ of the problem already.
That being said data access is the main bottleneck these days imo which is why I think a CAP-Theorem question is mendatory for pretty much every job interview.

If anyone is a half decent programmer and wants to make > $90K a year working from home on your own schedule in basically javascript for Medical Research Back Office software PM me :-P

Warning: Its an awful / proprietary environment but it pays well!

Where is that its a trap meme?

I have always wondered what a job like that would be like. How many hours a week do your developers work?

I completely agree. Unfortunately, modeling complex applications in FP languages is really complicated. I think we're going to see some interesting architectures arise using combinations of OO languages and FP.

Sounds snobby. Python is a very useful programming language.

This is going to break your heart. Twitter is actually still using Rails quite a bit. Although mostly for templating and lightweight things these days. The "big" switch from Rails came when they decided all future services, and existing services that are heavily used such as the main firehose/streaming, will be written in Java from now on.

Regardless, it always amuses me that everyone uses Twitter as the example not to use Rails. It was a poorly written Rails app and let's be honest, we could all dream of having a tenth of the traffic that Twitter had back when Rails became the problem for them. There are plenty of companies using Rails that are doing more than fine. And most of the issues Twitter had weren't directly Rails' fault. Application Servers for Ruby have come a long way since then. Unicorn, Thin, and Puma are much, much better than Mongrel and WEBrick were.

Correct, these days Matz mostly writes C because he's continuing to grow the Ruby language. Aside from that though he uses Ruby any chance he gets. He designed it to be an "80%" language.

Solid point. Pointing out how a hugely successful company out grew X is never a good argument against choosing X.

Think about this statement next time you use a cell phone. I think you're stating this simply because you find FP hard, not because of any inherent failure of FP. SICP is a book designed to teach you how to think about building larger systems and it is written with a strong slant to FP. Scala and Clojure are written as FP languages because people in the real world figured out mutation in particular is really bad for complex systems.

There is some interesting BS code in the LC thread from Code Academy. Udacity was summarily bashed for showing students how to write awful Python code despite this code coming from Peter Norvig, and even MIT OCW had this slop thrown at it. What I'm saying is that I think Python's sudden and quick rise as a teaching language is ultimately going to hurt the language in the long run, sort of the state of PHP now. Basically we are going to end up with a ton of people with 6 weeks of education saying they can program in general and Python specifically.

This sort of red-herring argument isn't going to work with me for three reasons:

1- I never once bashed into languages.
2- I did not say that RoR is a lesser language because of Twitter's switch. As a counter-example, Reddit was originally in Common Lisp and it is now in Django, and that is perfectly okay by me in case you get the impression that I dislike Python from the above paragraph.
3- Your statement has absolutely nothing to with the point I was making.


What proof do you have that the code-base was bad? Did you used to work at Twitter? Did Twitter open-source all the code they used? Is Jack Dorsey's github page an eyesore of BS code? After 20 years of writing code is his code-writing skills terrible?

The common complaint against RoR in particular, at least what I've read, is that many times people just gun up RoR without knowing how to write Ruby at all, which I agree is really bad. I spoke to one web dev firm here in Los Angeles and they switched away from RoR because it got way to difficult to find people who could program at all. Yeah, they could scaffold and install scripts, but anytime they were faced with an issue that wasn't prefab, they couldn't figure out how to work through it. This, I fear, is the direction Python is heading into and I hope you agree with me that this is a very bad state of affairs no matter what line you draw in language wars.

By using Scala in your example you've proven my point about hybrid OO and FP. Thank you

I don't really see how this hurts the Python language. Are there features being added to Python that are only added for the benefit of early/poor programmers and that hurt the language?

I don't follow the politics behind Python very closely but my understanding is that Python is a particularly bad example since Guido still rules on major language changes.

This depends on your definition of OO. I use a very simple definition: A model used to control mutation.

Clojure, Erlang, Scala, et. al. all have this definition of OO included, but that makes them no more classically OO than Haskell (which has mutation as well). Of course there has to be some mutability constructs available in all languages, this is a huge contrast to languages that actively discourage FP code. Just because FP languages all have mutation available doesn't mean that using mutation for large concurrent systems in encouraged and holding up examples of mutation in these languages shouldn't be misused as vindication. Yes, these ideas fair slightly better because they don't default to mutation and yes, you can add "OO" to them if you feel like it. Do whatever you want with them. That is why Candybar called them more expressive than other languages.

this is overly simplistic to say the least

Yes, it is because that is the only piece of the oo vs fp area that matters. FP still has polymorphism, inheritance, and other concept, including mutation that are shares with oo. There is no full blown dismissal of oo here so saying there is a cross pollination isn't debatable or particularly revealing.

i don't what you're talking about, i wasn't weighing in on your debate with richi, i just noticed that comment and thought it was missing the point of OO.

Yes, and you are also aware that I'm about AA far removed from OO guru as one can be.

disagree again

I run a consultancy firm so hours are flexible but there's pretty much unlimited budget currently so can work anywhere from 20-60 hours a week at $50-70+ / hr for the easiest development ever.