I hear this comment frequently regarding Linux. I think this points out
what is, in my mind, the major weakness in Linux that keeps most
non-technical users from trying to use it. If a non-technical user has to
know more than he/she does about their computer in order to use a piece of
software, most users will find another way to get the job done. An
unreliable solution such as M$-Win will seem preferable to a reliable
solution like Linux simply because the user will, in most cases, be able to
get it to work, most of the time, with minimal help. And thats usually
preferable to getting lots of help to have a more reliable system.
The big advantage that M$ has in this area is their vast database of
hardware information and an installer that can usually identify and
configure the OS installation without the user knowing much about whats in
their box. Sure, it doesn't work all of the time. But it does work most of
the time and thats good enough for most non-technical users.
Microsoft can easily collect configuration information from hardware
vendors because of its dominant position in the marketplace. Linux
distributors are at a major disadvantage. They don't have the size or
resources to collect and put to use the vast amount of info that it would
take to make Linux installation fool-proof. For something on this scale to
happen in the Linux community the information will have to come from the
community itself.
Maybe it would make sense to have an on-line database of configuration
information that a Linux installation program would reference when setting
up a new machine. All Linux users and distributors would then have access
to the same data. As the database grows so does the ease of Linux
installation.
As Linux developers and users create new configurations for new hardware,
they would add their collected wisdom to the database such that install
programs would be able to use the information to identify the same hardware
in other systems and set up the OS as required. Given that the data is
stored on-line, all install programs could have access to the latest and
greatest configurations.
I am new to the Linux world so I don't know the true scope of a project
like this but it sounds substantial. I could see there being a need for
some kind of data collection program that could collect configuration
information from a users new system and add it to the database. Of course
information would also have to come from Linux developers too, especially
driver developers.
There would be problems with keeping the database free of bogus
information. Seems like this would be a real challenge in an environment
where lots of developers and users can post configurations to the database.
Maintaining a clean configuration database is easier to address for a big,
centralized operation like M$. A lot harder to address for a loosely knit
group of developers such as those working on Linux.
Maybe a voting scheme could work. The installer programs could look at the
database and compare the various configurations for the hardware it sees on
the local machine with the entries in the database. If there are more than
one entry for the same hardware then the installer could choose the
configuration that appears in the database most frequently. Maybe the
installer could make the decision based on where the config came from. For
example a config from a hardware vendor, or the author of the driver, may
be thought of as more reliable than a config from an unknown user.
If Linux is to seriously compete with M$ on the desktop, something will
have to be done in order to ease the burden of installation and
configuration for the non-technical user community. No small task. Of
course Linux was no small task either.
Just a few thoughts from someone new to Linux....
Dave
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/