[svlug] ubuntu 14.04 where are the wifi tables stored?

Steve Litt slitt at troubleshooters.com
Mon Jan 19 09:44:24 PST 2015

On Sun, 18 Jan 2015 11:05:19 +0100
Ivan Sergio Borgonovo <mail at webthatworks.it> wrote:

> Linux on the desktop is a fairy tale. It's a miracle we have so
> functional graphic environment considering there are very few money
> you can earn from writing a DE for Linux.

Precisely. I slept a hundred years, let down my hair from the tower
window to the ground, got kissed by a prince, and now I'm using Openbox
with Inkscape and Sigil!


OK, seriously, I've pondered the whole "why do they give it away"
question, and here are some of the things I've come up with, at least
for small projects like Openbox or runit:


I never got paid a cent to write VimOutliner. But it's software,
integral to my business, that gives me a competitive advantage. And the
VimOutliner I use today, which has been improved by generations of
developers, is much more useful than the VimOutliner I cobbled together
from Perl and Vim in 2001. So I can write books faster, plan better,
and have a more profitable business.

I'll make one other point, using VimOutliner (VO) as an example. Red
Hat wasn't paying us. We didn't need to justify our progress to an
employer or patron. We all had day jobs, our time was scarce, so when
we added features, we did it in the simplest possible way. Right from
the start.

Needing an outliner, I could have written something in Python Tk. But I
didn't have the time: Red Hat wasn't paying me. So I took Vim, used a
few Perl scripts to make its folding feature into outline
expand/collapse, and a few days later I had an outliner, *and then I
stopped developing and started using*. 

As other people began developing VO, the only things that went into VO
was a scratch to the developer's itch. No special code for Retina
Displays, no Socket Activation, no Desktop Environment bindings. As a
matter of fact, two features we all wanted, clones and collaborative
outlining, were never done. They were just too difficult. And therein
is found the true value of small Free Software.

Imagine, for a second, that IBM was paying Noel, Matej and myself to
develop VimOutliner. You bet we'd have put in clones and collaborative
outlining, even though both run contrary to the very workings of Vim.
We'd have quadrupled the code base doing it, probably built a house of
cards architecture, but we'd have done it, because that was our
paycheck. And forever after, all VO users would have had to live with
the bugs we introduced when we added those two features. Even those not
using those two features might have been bitten by those bugs.

All the time, I hear people gripe that corporations use Free Software
but don't contribute back, either with software or money. My viewpoint
is that, just by using Free Software and introducing it to their
employees, they contribute plenty.

But beyond that, corporate money can corrupt Free Software. It doesn't
have to, but it can. A salary generates an expectation of continuous
improvement, the paid developers comply in order to keep their jobs,
and all too often the "improvements" are setbacks. Gnome2 becomes
Gnome3. Kmail becomes Kmail2. A PID1 becomes an "OS Toolkit". These
things don't happen with unpaid development.

Hey, let's not run from the truth: Unpaid apps usually have that "Army
surplus" look and feel to them, especially when compared to the
sanitized Windows UI and that fine art masterpiece called OS/x. But
unpaid apps are simple, they usually have few bugs, they often have few
dependencies, and they're adaptable enough that you can change them to
solve your exact problem.

Bringing it back to the original topic, it *is* surprising that KDE and
Gnome got developed, and these certainly wouldn't have been developed
without a lot of developers who were paid to work on them. But if KDE
and Gnome were to instantly vanish, things like Openbox, dwm, dmenu, i3,
jwm and the like would keep being made by people scratching their
itches, and those are more than sufficient to run dia and Inkscape, and
perhaps even Gimp if Gimp doesn't get too saddled with unnecessary


Steve Litt                *  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance

More information about the svlug mailing list