|
Why Linux Cannot Rule the World, Yet
Krishna Kumar
Wednesday, March 08, 2006
World domination, even if lightly stated is one of the
stated ambitions of Linux. And in recent years, Linux has made enough progress
for many to think that world domination is indeed within reach.
I beg to differ
I have worked on Linux for about ten years now. Ok. I have supervised work
on Linux for about ten years now. My teams here have worked on them, like most
people using Linux will never have. They have installed Linux on all sorts of
hardware, and made it work with other operating systems. We have even rolled out
our own distribution year after year, and believe me this last one, rolling out
the distribution is no easy job.
 |
| Krishna Kumar, Editor |
One of the promises of OpenSource is that the code is there
for you to see, and modify. And that is easier said than done. Consider the
small example of Anaconda, the graphical installer used by most rpm based
distributions. Last year, we struggled to add an extra screen to the process,
and some how succeeded at the last minute. This year we tried and tried and
tried to change the font size in the Anaconda screens, and to centre align the
menu text; and failed miserably on both counts. We poured over every line of the
code, Googled all over the place and delved deep into libraries, and still
failed to find a way to make this simple change.
Why?
Because Anaconda is developed by many different developers with completely
different ideas and capabilities, but with out a common base framework and
without proper documentation.
Every time a new version of the kernel or even a library
comes out, more than half applications one is using refuses to work and there is
no knowing if and when a new version is coming out. I have seen developers
answer-when I get the time-and can understand their compulsions. But then
don't expect me to use your application and help you achieve world domination.
Before you begin to think otherwise, I am not rubbishing
Linux or writing it off. I am not saying that it has no place in the larger
scheme of things. What I am saying is that Linux had reached a certain level by
following its current irreverent, devil may care, I am God attitude.
But for it to move on from there, there is a need to bring
more method into the madness. There is a need to ensure that there is enough
documentation and understanding of the core processes of the software and how to
get it working. There is a need to bring in a minimum level of assuredness to
the user that the application will be there and will continue to be updated; and
that the updating will be coordinated with upgrades in major elements like the
kernel.
Some of you may be thinking that I am recommending the
corporatization or commercialization of Linux and OpenSource. I am not. I am
advocating more responsibility towards the user. More assurance to him that the
application will continue to exist and will be upgraded as other elements of the
Linux ecosystem get upgraded.
Today, all software vendors are open about the fact that
they are trying to learn from, and incorporate good practices, from the
OpenSource movement. There is no reason why the OpenSource movements should not
learn a few lessons from them, where necessary from them where necessary.
Page(s) 1
|
