Comments about Linux/Ubuntu from a former MS-programmer

[Eric Dunbar]

On 10/04/06, Derek Broughton <news at pointerstop.ca> wrote:
> > that Nautilus doesn't play nicely with  the network,
>
> I didn't think Nautilus was supposed to play nicely with the network.
> That's what Konqueror is for. :-)

What if you hate Konq even more than Nautilus? Seriously though, Konq
and Nautilus are both supposed to be the GUI file browsers for their
respective environments. Nautilus has a hell of a hard time dealing
with my SMB file server. Mac OS X Finder OTOH plays perfectly with it
every time.

But, as with a lot of things in OSS, "give it time". Eventually
someone might stumble upon this as a good idea and go after it.

> > that there are numerous user interface flaws and that
> > apps don't play nice with each other.
>
> As opposed to Windows where there are _no_ UI flaws? :-)  On the whole, I
> find that the interface in Kubuntu is _better_ than in Windows, but while
> MS can enforce a consistent UI, at least for their own programs, it's
> difficult to do the same in the Linux world.  Generally, I just find that
> if there's an app that doesn't play nicely with the rest of the tools,
> there's a slightly different program that does.

As mentioned, I despise Windows. It's a poorly designed interface and
doesn't do a great job of implementing UI research. But, it is
certainly a smoother ride than GNOME or KDE (though, to their credit,
GNOME and KDE have made great strides in recent years). Mac OS X,
though certainly a far cry from perfect (in many ways having regressed
from its Mac OS 9 heritage), is tragically the "best of breed" (and,
that's not saying much).

I think the next revolution in GUI might come from the OSS world, but,
as long as people continue to copy Windows (and, sometimes Mac) we're
not going to see that happen. Of course, it's quite possible that
Apple will have something up their sleeve if they don't completely
drop Mac OS X and become a clone manufacturer ;-(

> > There are obviously people who can argue that Linux is ready for the
> > prime time and works perfectly for them on all fronts until they're
> > blue in the face, but, just like in a religious debate, whether or not
> > you believe isn't going to make something real!
>
> otoh, your assertion is no more valid.  If I argue that Linux is ready for
> prime time and works perfectly for me on all fronts (well, I wouldn't
> really do that...) you're in no position to deny my experience.  I no
> longer have any software I must run on Windows.  The flaws in such things
> as OpenOffice are acceptable for a generally superior user experience.
> YMM, of course, V :-)

The YMMV is exactly what's the problem and why my assertion _is_ more
valid. On Windows and Mac OS X your kilometerage does not vary
particularly much -- there is a consistent GUI for most apps, and,
they do play ball with each other quite nicely (at least, on Mac OS X
they generally do).

On Linux, you're taking your life into your own hands if you don't
have someone knowledgeable supporting your ass.

> > My diagnosis is that the problem with Linux is that it doesn't have
> > anyone pushing to get the newbie bugs fixed first.
>
> _If_ that's the problem, it's never going to go away.  Linux (and OSS in
> general) is evolutionary.  Good traits are kept, bad traits go away.
> Nobody is ever going to direct more than a small portion of development in
> any organized way.  That's both what makes Linux great ... and doesn't.

The social Darwinism inherent in that statement doesn't quite work.
Bad traits don't go away in the world of programming -- 15 years later
we still have Windows, 22 years later we still have a Mac-only file
system.

> > it sometimes appears that bugs get fixed when the dev decides to work
> > on it, not because an important user scenario is broken.
>
> Not "sometimes".  Always.  Arguing against the methodology is pointless.
> There's no other way that OSS development _can_ work.

I don't think that's true. With advocates or voting systems you can
show what's important and what's not. To a certain extent, I'm sure
that it already happens, but a formal and widely known system would
certainly help the process along. I think there are many developers
who are interested in making Linux accessible -- and, to a certain
extent they are succeeding.