Doing our Linux Final project, I got to experience a little bit of the Linux design philosophy. Our project, basically followed a building blocks structure where there were python scripts written to run in the console level and there was a GUI interface built in Qt that sat on top of that. This is a typical convention for many Unix applications and I wanted to share the various problems and aspects I experienced. For one, since Linux is based so much on collaboration, I noticed it was quite difficult to blindly have a developer working on the console level and then blindly have another developer working on the GUI. More often than not, the developer working on the GUI will need a working console application that he can play with and see how it works so he can implement it in his GUI. However, I was curious to see if there were any projects that have a strict protocol that the developers of the console and GUI both strictly oblige to. For example, if there was a protocol that specified for these specific inputs, you would get these outputs formatted in this way. With that knowledge, both developers can work hand in hand at the same time to get their respective parts working in parallel. I realized that this approach was far more streamlined working on our group project. Tyler and I worked on the GUI and Sasha and Sungha worked on the console python code. Rather than waiting for them to finish their code, we established what kind of output we were expecting from the console and how it should be formatted so that we had a working template and framework setup while they finished their code. This allowed us to have a somewhat working interface that we can fine tune once the console application had been completed.

Researching into how other developers and things handle this method of establishing protocols, I came across the IMAP protocol. This to me seems to be the most standardized and widespread protocol used to this day and it is used for sending and fetching emails. I was intrigued by how millions of various applications, interfaces, web servers, and consoles conformed to this protocol when it came to the implementation of email services. If only there were protocols such as this that can be established for various types of data so that developers can collaborate much more efficiently and quickly.

Today I met with my group for my final project and setup a github to use for our project. I noticed a lot of nice features for github that allows us to collaborate effectively. The way we distributed the tasks was that we had various functions each of us is going to write and then we were going to put the functions together in a main file that would run. This method of organization worked out well for github because it allowed each of us to work on a specific file and commit changes to that file. While I like how user’s can commit changes and upload it to github, I feel that it is a little old fashioned. What if github was like dropbox and it automatically synced all of the changes and automatically kept track of change history each time a file was saved. I understand that in github it allows users to only commit once they have a good working piece of code but it is annoying when multiple users are using the same repository and making changes. Each time the users all have to update or pull their repository to update it. Rather, if it was structured like dropbox in which the files get automatically synced, it would be much more user friendly.

After using Ubuntu quite regularly for quite a long time now, I have grown accustom to the  fast boot up, fast processing, and great customization features. I dread the Sunday nights that I choose to boot into Windows 7 to work on my AC Circuit Analysis Lab assignments. Ubuntu without a doubt has a very minimized package set that makes it as quick as possible. Windows 7 on the other hand, is bogged down with a bunch of unnecessary bloat ware and applications. Ubuntu on the other hand, is stripped down of all of this and I notice this via the significant performance difference.

However, after some research, I have noticed that Ubuntu also has some “bloatware.” What does that mean? It means that there can be an OS that can be built that gives an even bigger performance boost than Ubuntu. We do not use ALL of the packages that come with Ubuntu and I think that it is an important part of Unix to be able to have an OS that is as customized and stripped down as possible for the user. As a result, I think Arch Linux is the way to go. Over winter break, I plan on getting a new SSD, partitioning it, having Windows 8 on one partition, and Arch Linux on the other. I think by installing Arch Linux as a core Linux installation and building up your Linux environment package by package, one can acquire a very intricate grasp of Unix and appreciate the platform much more. I hope to do this over winter break and see if I can really push the limits of performance for my computer.