Maximum PC

Bought Algorithms in a Nutshell, The Ruby Programming Language, and Erlang Programming.

I'm under the impression that there are certain aspects of computer science that aren't taught or taught well. Having written so much C# code, I'm making the move towards Open Source this year. I believe that Open Source has solved some significant issues in computing a full decade ahead of Microsoft (IMHO).

Algorithms in a Nutshell: I believe that I was gypped in my algorithms course. We had spent so much time on the mathematics behind the algorithm that we never really became all that familiar with the algorithms themselves. I'm devoting time to relearning these algorithms for my own sake, I believe that having these under my belt will benefit me greatly. These time tested algorithms are the basis for a lot of the already-abstracted things in languages. For example, sorting is a huge topic and different frameworks will use any of these time-tested sorting algorithms.

Plus it looks good in an interview. I'm going to add algorithms to my repertoire of interview questions. Fundamentally speaking, good programmers should have a solid command of them, languages not withstanding.

Erland Programming: I wrote about this in an earlier post. Erlang is a highly distributed, highly concurrent, soft-real-time programming language. Erricson had the idea years ago that a many applications will utilize concurrent, scalable and distributed properties. Working for a hospital, even healthcare utilizes many of these technologies but they're often coded in languages that hinder true concurrent, scalable and distributed architectures.

While .NET may be concurrent/parallelized, the implementation of how it performs distribution, concurrency and scaling is often obscure. Has anyone tried writing a distributed anything in .NET? If so, it's a bit of a pain! Parallelism and multi-core functionality didn't really arrive until .NET 4.0 (at least not implemented in the framework). Erlang won my respect (and heart), so you may wonder: where would I implement Erlang?

In a hospital setting, it can be used as a scalable, distributed data repository. Architecturally, I would only have to worry about the actual implementation and let the language do its job. I would totally use it to build interface engines between non-compatible applications (w/out using ORB's). I believe that Erlang can do it and do it well.

Ruby Programming: I don't believe that one should learn a language through a framework. Currently, I understand Ruby from a Rail's perspective. I see Ruby through Rails, I want to understand why Ruby is Ruby. Rails is a wonderful framework, but I want the power of Ruby in my hands. This should empower me to use frameworks like Sinatra and Rack. Ruby's a powerful language and if the community focuses on making Ruby development simpler and easier, it can supplant the LAMP and .NET stacks easily.

I'd like to hear your reasons on this. Need a good discussion here!

I am not so sure that F/OSS really has "solved" anything that far ahead of MS.

It is good for you to expand though, I wish I had more time to myself! However from what I have seen, I really believe that the biggest problem with MS is not MS technology but the people whom implement it!

You can have shit code just about anywhere and in any language.

It stems from this article. The CS Dept in my uni focused heavily in the mathematics department, which isn't necessarily a problem. When I got out of college, I struggled with working as a software engineer partly not because I didn't know the math, mostly because I didn't realize that 90% of the coding I do is mostly consuming an API or a framework. Granted this takes time, but all those projects in school writing binary search algos didn't really prepare me well for the real world. Often, academia is so obscure and distant from the real world, it took me a good year or two to get over the hurdle. I found out quickly that I never had to write a Bubble Sort or Binary Search algorithm after uni, though I was tasked to solve more complex tasks like finding ways to make disparate systems talk...something I had little prep from my Uni.



I think they have. Erlang was designed to be a scalable and distributed language. Microsoft doesn't even have patterns for that and .NET certainly isn't designed for it. Erlang's also multi-core, multi-processor and can do soft real-time execution...well ahead of C# and .NET. Prolog solved the problem of building logic; Prolog's designers answered the question of "how do we create a language that builds the logic for you?" Prolog's design is more data oriented based on facts and data, then building out the logic to determine the result. You can do this .NET, but it's really obscure and difficult. Web Dev is the biggest piece where MS is beginning to lag in; jQuery, Prototype, Backbone.js, node.js...Rails is taking over market segments that PHP and ASP.NET once ruled. The F/OSS community is now bringing true MVC patterns by utilizing the prototypical features of Javascript and create true, reusable services through REST.



Hehehe, true! I'm taking time in the evening after dinner, before working out. I'm hoping to work on an ORM Mapper that's not as complex or sophisticated as Entity Framework, but something that can handle relationships (Dapper.net doesn't handle relationships). If you want to take part in building it, let me know, I can github/codeplex it. Ideally, it should be simple...maybe I'll start a new thread about it?