Debian: contempt for "end user" values has to stop!"

David Gerard dgerard at gmail.com
Fri Aug 21 12:09:12 BST 2009


2009/8/21 Derek Broughton <derek at pointerstop.ca>:
> David Gerard wrote:

>> A volunteer will work ten times as hard as any employee ... doing what
>> they damn well please.
>> So the necessary management skill is cat-herding: you need to work out
>> the local value of tuna.

> LOL.  I think that says what I've been trying to, perfectly.


I've herded volunteers in various contexts (a fanzine, a private
school fundraiser, a student newspaper, Wikipedia) for nearly 25 years
now. It's a hard-won lesson to learn!

In the context of Linux distros: computers are already annoying
enough. People who have a system that does what they want will not
change it because someone else wants to beat Windows. "You should stop
using KDE because we have to beat Windows." "Er, no. Bugger off."

So people who want to beat Windows have to work out a way to make the
variety of Linux into an advantage for that goal.

Suggestions: a lot of the problem with Windows is that one size does
not in fact fit all - it's a morass of bad hacks and compromises they
now have to maintain forever. So variety and fitness for purpose is an
important advantage - Ubuntu is the most common desktop Linux, Red Hat
is by far the most common server.

(That said, there's two jobs I recently applied for *because* they
were running Debian. Not that Red Hat/CentOS isn't a perfectly good
Unix-like operating system that does pretty much anything you'd want a
Unix-like to do - but that they were using Debian showed they'd
*thought about* what OS to use, rather than just using a Red Hat
variant because everyone does and it's something you won't get fired
for. Of course, I put forward my experience running and unbuggering
Ubuntu alphas as relevant ;-) )

Of course, the other alternative is Mac OS X, which is a perfectly
good Unix(tm) and is much more one-size-fits-most. And its design
isn't hacky. (Mostly.)


- d.



More information about the sounder mailing list