[erlang-questions] Coming Back (maybe improving lists:reverse/1)

Ivan Carmenates Garcia <>
Thu Oct 8 01:43:09 CEST 2015


Hi Cian, well I have no long list, short list actually, but what I want is millions of repentances of the same algorithm. i.e. for concurrency of millions of users. Lol well maybe I was a little bit extreme here.

I treasure Erlang as the best language ever and the future language, multicore, concurrency, etc. That's why maybe sometime I could get so extreme!. For example I wrote this about Erlang a few days ago and sent it to a person to defend it against the world and get a job, please correct me if I am wrong in some aspects.

The future is multicore, and each day more computers with more amount of cpu cores and the world is more concurrent. We choose Erlang because it is the language of the future so it is already ready for multicore programming and there is no other language like it in this aspect.

The majority of programmers rather easy and comfortable languages like Php or Pyhton, because of its beautiful syntax, but this languages are not ready for the future, they are sequentially and has no clue about what multicore programming is or it is very hard to implement.

Python and Php for example are interpreted languages and because of it they are very slow so we have to spent more resources in servers and hardware to get more reach. That is why facebook had to made its own version of php compiling it to c++ looking for performance, they had to do it because a huge of amount of programmers that php has because the necessity of gain money quickly. However facebook choose to implement its chat system using Erlang, they spend two years of research to converge in that choice. Python for example is a better language than Php because it is well structured and more powerful yet is a script language similar to JavaScript and its better end is for make plugins and utilities libraries.

Erlang is growing up because of its power to handle millions of concurrent process without any difficulty and very easy to implement and that so to build entire distributed systems with huge reliance. It is not only the way of programming easy, it is to do it so a little more complex however not with more effort and adopting the new tendency that is distributed functional, concurrency and fail tolerant oriented programming.

A simple example is: The Oxford University of England does not had any postgraduate courses last year in Informatics Sciences and they open one this year and guess what the topic of the course is: "Modern programming languages, including declarative, logical and parallel paradigms, functional programming, concurrency, distributed systems.".
...

Well sorry about the speech lol and the translation I just made, I wrote it originally in Spanish.

Best regards,
Ivan (son of Gilberio).



-----Original Message-----
From: Cian Synnott [mailto:] 
Sent: Wednesday, October 7, 2015 3:48 PM
To: Ivan Carmenates Garcia
Cc: 
Subject: Re: [erlang-questions] Coming Back (maybe improving lists:reverse/1)

Hi Ivan,

On Wed, Oct 7, 2015 at 7:39 PM, Ivan Carmenates Garcia <> wrote:
> lists module, and doing some recursive functions, it is usually a 
> problem that we need to do lists:reverse at the end of the algorism to 
> get the data in the right order.
>
Are you sure that it's a problem, e.g. do you have very long lists, or measurements that demonstrate an issue?

See http://emauton.org/2015/01/25/lists:reverse-1-performance-in-erlang/
for a brief note on why you probably don't need to worry about this.
:o)

> So I can imagine Erlang implement lists using double linked lists for 
> obvious purposes,
>
They're singly-linked - this is why we tend to build them in "reverse order" in the first place.

Cian



More information about the erlang-questions mailing list