This page is meant to provide additional thoughts to [WinningTheDesktop]. Although I like the ideas presented there, I think this is already the second step - winning the Desktop starts with anticipating the user's WAY of doing things and with providing an alternative to existing systems. And when I say user, I mean the 'average Windows user'.
I don't want to convince anybody of anything - the following is just my personal opinion and some kind of 'mind dump'.
Winning the Desktop - hum? Already there.
<quote>Linux will never be a one-click-easy-to-use OS. It's a system from developers for developers. We can already do everything an XP User can.</quote> If you agree with this statement this probably isn't for you. Software is a means to get a job done and not some self-sufficient sort of thing. Software should be free for everyone - not only for geeks.
Winning the Desktop I - Where's the problem?
Sticky Information. The Knowledge which is necessary to create a product or to develop software can be categorized in:
- Knowledge of the purpose / needed features / usability of a certain product / software (provided or generated by the USER)
- Knowledge of the technology to produce the product / software ('Know How') (provided or generated by the manufacturer / developer)
The problem is that both kinds of information are 'sticky' which means that they are not easily transferable from one person or group to another (for a thourogh explanation you should read http://web.mit.edu/evhippel/www/democ.htm ). While the traditional (commercial) approach to this problem is accumulation of production knowledge and observation of the market to obtain the necessary information from the user, open source incorporates the whole knowledge in a single person ('user-developer'). At least this is true for software targeted at technology-affine users... Unfortunately this is not the case when it comes to a click-and-drop-easy-to-configure Desktop, because the knowledge of the wanted features is that of a technically skilled person and not Joe User's.
Winning the Desktop II - but Open Source means Work in Progress - we're getting there
Not so sure. Almost every 'Desktop-Linux' Distribution on this planet tries to make Linux *LOOK* like Windows or OS X (that of course wouldn't be too bad...) without going radical ways to simplify the system itself. Possibly there is a trade-off - Can a Linux distribution do both? Satisfy the needs of developer AND Joe User at the same time? I don't know. It is like trying to make a 10 ton truck look like a Ferrari and as easy to drive as a bike.
Progress means evolution.
Evolution in a Software development sense means:
- Selection - Survival of the best ideas / projects
- Mutation - creation of different approaches to a certain problem
- Crossover - combination of ideas from different projects
Open Source is especially superior to commercial software development in finding different solutions to a common problem and in insuring that the 'basics are done right'. The combination of different projects and the merging of ideas is also a strength of Open Source development. Selection is the part where Open Source development is really poor at. Because there is no 'natural death' of a 'bad' project (equivalent to a to small market share or the end of the life cycle of a commercial product) outgunned projects survive and absorb valuable resources. I don't want to start a disussion here why Windows then still exists - maybe that's a market failure caused by network externalities http://www.utdallas.edu/~liebowit/palgrave/network.html.
Winning the Desktop III - Choice is the essence of Open Source and what about *equally* good projects?
Assume the following: Joe User wants to listen to some MP3 located on another (Windows) machine while writing a letter and burning a cd on his recently setup 'Desktop Linux' box. Like so:
- What was the name of the MP3 Player again?
- O.K. found it, but where is my win share?
- (Somehow figured it out, after setting up Synaptic, installing and configuring ntfs and samba which is absolutely unprobable) but why can't I play MP3s?
- O.K. legal issues... figured it out...
- Where's a program to burn CDs?
- (Google said K3B, very intuitive name by the way, is fine)
- (after finding out that downloading software and double-clicking to install it is kind of impossible, using Synaptic)
- O.K., Starting K3B... Why the heck does this look so different??? And what is KDE?
- O.K., there is OO - that looks familiar - it's like Office! (*writing letter*)- now saving the document...
- HUM???? /etc /root /home /var /dev what the ....?
Get the idea? By the way Joe is Nobel Prize winner but his time is scarce and configuring a system is a waste of time if you already know how to do all this on Windows. And I just touched the surface - dependency hell, config scripts, runlevels, compiling from source, sudo, xhost +, SWAT, CUPS,...
IMO a distribution is all about selection *NOT* choice. If you say Gnome it's Gnome. If you say gnome-conf it's gnome-conf,... And then it is about progression - choose one direction and run.
Of course there are equally good projects for one purpose - but Joe does absolutely not care - he wants the one which doesn't get in his way - no matter if there is a project XYZ which has another feature ABC.
Winning the Desktop IV - is this all about critizing the very nature of Open Source?
No way! I believe that OS is a very good thing - not only because it produces superior software but also because it is about freedom (as in free speech AND in free beer). What I'm trying to say is that because of the very nature of OS it won't be easy to arrive at the Desktop and I would love to see especially Ubuntu on the Desktop, because many things here are already done right.
Winning the Desktop V - HowTo?
Winning the Desktop cannot be about cloning Windows. Period. Even if Ubuntu succeeds in being as easy to use as Windows I doubt there would be a noticeable impact on the MS market share. The only solution is IMHO to actually be *BETTER* than Windows in terms of usability.
I very much agree with this statement: http://www.eweek.com/article2/0,1759,1824810,00.asp The time is now and the concepts are there.
- Fetch the user from where he is!
- Allow him/her to do get a job done with minimum hassle
- Take a look around and incorporate the most science-fiction projects (and only those!)
Provide a reliable platform (very important for commercial software!) as suggested here: http://autopackage.org/faq.html#5_1
Easy isn't it?
Some uncommented and unordered links