Why The Linux Desktop Matters More Than Ever, This Time?

This article is talking about Ubuntu 12.04 aka Precise Pangolin

If we look back at failures of Linux on Desktops to this day and analyze the technical part, it was due to hardware support and the software quality (GUI) that general computer users could use.

This time however with Intel having already released the graphics drivers (after a disappointing year with sandy bridge) even for upcoming hardware aka Ivy bridge, kernel developers and open source graphics hackers can sit back and relax (not really). With new xinput (multi-touch), multi-monitor support and some software that matters appearing in the Ubuntu Software Center.

The stage is set for Canonical to get the user experience right for the normal user. Here normal user is the one that may come from any platform, or may be new to computers. Normal user is going to buy Ubuntu laptops, tablets or burn a DVD or install ubuntu via flash drive within half an hour.

Within that half hour, if a user can boot from usb and get his/her mov or mpeg file playing then we can say canonical did their job well. Besides having the multimedia file work he/she should have no issues at all (and I mean any at all), that would make them frustrated, and the quality of software that goes into that hardware should bring no issues (graphics, power management, networking and security) at all. It should save user's preferences (brightness, history, or not save history if user chooses to do so).

The bottom line is no update should break anything or bring any major changes to the user experience. Software updates should be configurable to update only when user chooses (daily, monthly, manually etc). Important security updates can have little more granted, but not to the level where it interferes with the user's experience (e.g dirty and noisy notifications that we always tried to avoid).

The sound experience should not break, themes included should be consistent. Mouse cursors, if included, should be consistent around the whole Desktop. Music applications like rhythmbox must have equalizers with presets and should sound better when the user chooses them. Calendar application must return back to Ubuntu. And Ubuntu should get a new icon theme. In my opinion that is the base line for ubuntu 12.04.

If they get it right this time or even complete these major tasks, will not get a half baked ubuntu experience and we can recommend ubuntu to anyone. I use ubuntu on my office and home. One of the junior developers, who has never used Linux before started to use it. He could use it because I left the computer he started using and switched to new one. I had configured everything I usually do after installing Ubuntu.

Stuff like sound issues, wireless network, proprietary drivers and installing basic software used for development. His expertise on Linux is very low. i.e. He does not care about the terminal (calls it dos) and uses it only to run mvn and svn. He can connect DSL and wireless internet and run other development tools. I can say with confidence, anyone can use a well configured Linux. But the gap between unusable Linux and unusable Linux is very big.

I had to remove annoying update notifications (I don't want his Ubuntu to stop function, for no apparent reasons), install wireless drivers and setup some kernel parameters like fbc and aspm. With new kernel implementing aspm and stuffs, the only issue that remains is the wireless.

"If you don't have an internet connection, how can you know why the internet is not working" - South Park.

Over the years and recently via direct feedback tests and bug list over launchpad, Canonical has gathered almost all the problems users have. Even if the 'upstream' keeps on complaining about Canonical not participating on code development (they might be right), but the user feedback Canonical collects is priceless even for those upstream developers.

If we go back 5 years and see to today, we all know we have come far in a positive way to strengthen the quality of Linux software, user experience, reliability and more which Canonical has a great share on. At the end of the day developers want their code to be useful to others. The quality of our code can only be measured via user's easiness to use the software and their productivity increase via that software.

I often find comments like "can do". "Can do" implies if there is a feature in the software, users can make certain changes on the settings to make that feature work. We Linux folks have wanted features so badly that we do not care about how an average computer user is going to use that software. Due to such configuration developers always become lazy to provide basic features and leave it open.

Most Linux software has always been heavily configurable via settings file. We always follow "Do one thing and do it well". As a result we have fine grained packages (all configurable), which are the basic components of all our complex software. That means even if we wisher in a library we might crumble the whole software city.

This is the reason why I like Ubuntu's decision to remove ccsm from software center. I like consistent look of unity desktop (even I might have my own opinion about what ubuntu should look like and behave). If I have to go to the machine city (eg in virtualbox), I can remove all GUI and load just the terminal interface or go arch way.

That "can do" is for people like me who cannot settle until they break Linux. I am against every Linux Desktop to look the same, unless they have some targeted software or users. I want Ubuntu to be Ubuntu and nothing else. If I don't like Ubuntu I might not use it or make configurations as I want, but Ubuntu must take the path of normal user centric Desktop, and provide features that would not break the user experience on any way.

The theme of ubuntu "Linux for human beings" is what it should be. Ubuntu for this release and until 14.04 must focus on usability rather than anything else. Any human being should be able to use it, and should have no previous knowledge of anything to use it. They must test from a user's prospective rather than a developers prospective.